Vega Admin Guide R85 v1.6 PDF
Vega Admin Guide R85 v1.6 PDF
Vega Admin Guide R85 v1.6 PDF
c127=127
From Release 7.5, the Vega can also apply a cause code mapping to cause codes received from
the (ISDN or RBS CAS) telephony interfaces. Cause code mapping tables are configurable
through the web browser using the advanced>show_cause_mapping menu or via the CLI
parameters:
[_advanced.incoming_cause_mapping.x]
name = <name> ; name par amet er f or sel f document at i on pur poses
c1=1 ; mappi ng f or cause code 1 ( by def aul t = 1)
c2=2 ; mappi ng f or cause code 2 ( by def aul t = 2) et c.
c127=127
Cause code mappings are set up by altering the cause code parameters away from the 1:1
relationship (c1=1, c2=2 etc.) which is the default configuration. If a call comes in with a
cleardown cause code of 2, for instance, then the Vega will look up parameter c2 and will pass on
the value that has been assigned to it as the cleadown cause code.
Each ISDN interface can be configured to map or not to map cause codes using:
[e1t1/bri.port.n.isdn]
i ncomi ng_cause_mappi ng_i ndex=x ; i ncomi ng mappi ng t abl e t o use
out goi ng_cause_mappi ng_i ndex=x ; out goi ng mappi ng t abl e t o use
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 197 - 6/2/2009
x defines the _advanced. cause_mappi ng. x mapping table to use. If x = 0 then no mapping
is performed.
The mapping table to use for each ISDN interface may be configured through the web browser
using:
e1t1/bri>Port Configuration Modify>e1t1/bri_configuration >ISDN Configuration>
cause_mapping
10.2.7 Bus master
The bus_mast er _pr i or i t y configuration parameter defines which trunk the Vega uses to
synchronise its internal clock.
The Vega receives a clock on ports configured as cl ock_mast er = 0 (Vega 400) and as nt =0
(Vega 50 BRI). The bus_mast er _pr i or i t y parameter should be configured to prioritise the
clock receiver trunks in the order that they should be used for synchronising the Vega internal
clock.
For further details on configuring bus master, see Information Note IN 03 ISDN Clocks
10.2.8 Vega 400 Bypass Relays
For more information on this feature refer to Information Note IN 44 Vega 400 Bypass
Relays
Some models of Vega 400 are fitted with fallback relays such that in the event of power failure or
intervention by maintenance personnel the E1T1 connections become metallically connected to a
second set of RJ 45 connectors.
The diagram below shows a typical install where the fallback relays could be in use:
The status of the ISDN fallback relays can determine whether SIP registration takes place on a
Vega 400 (models where ISDN fallback relays are fitted).
The slave Vega can be configured such that it will only transmit SIP REGISTER messages when
its DSLs become active. This would happen if the master Vega loses power, is upgraded or is
manually put into bypass mode.
Parameter:
si p. r eg_mode
Possible values:
nor mal Def aul t Exi st i ng behavi our , Vega wi l l r egi st er any
conf i gur ed SI P user s
on_I SDN_act i ve Vega wi l l onl y r egi st er user s when any DSL i s act i ve
pbx
MASTER SLAVE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 198 - 6/2/2009
10.2.9 Specific T1 configuration
10.2.9.1 T1 Line matching
For the Vega T1 product the transmit equalisation for the ISDN trunks needs to be configured.
This is achieved on a per trunk basis using:
[e1t1/bri.port.x]
t 1_t x_equal i zat i on=<t x_equ>
<t x_equ>can take the following values:
l hl bo0 (long haul line build out 0 dB)
l hl bo7_5 (long haul line build out -7.5 dB)
l hl bo15 (long haul line build out -15 dB)
l hl bo22_5 (long haul line build out -22.5 dB)
sh0_110 (short haul 0-110 ft.)
sh110_220 (short haul 110-220 ft.)
sh220_330 (short haul 220-330 ft.) - default setting
sh330_440 (short haul 330-440 ft.)
sh440_550 (short haul 440-550 ft.)
sh550_660 (short haul 550-660 ft.)
Long haul values are used where the distance between the Vega and the closest repeater or other
ISDN endpoint is greater than 660 feet. Short haul value lengths are the distance between the
Vega and the closest repeater or other ISDN endpoint.
The t 1_t x_equal i zat i on setting is only applicable in T1 mode
(t opol ogy=t 1); in E1 mode t 1_t x_equal i zat i on is ignored.
E1 systems have their own equalization setting e1_r x_shor t _haul
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 199 - 6/2/2009
10.2.9.1.1 Guidelines for configuring t1_tx_equalization:
For short haul the aim is to make sure that the shape of the waveform at the receiver is as perfect
as possible; changing the parameter alters the shape of the waveform generated by the Vega (to
compensate for the additional capacitance of longer lines). Match the parameter value to the line
lengths indicated in the above table. If the length is not known, then start using the value
sh220_330.
For long haul (>660 feet) the waveshape is not altered any further; the configuration parameter
affects the amplitude of the signal. The aim is to tune the transmit amplitude such that the receiver
receives a signal slightly above 36dB below the maximum signal strength (the 0dBm value). If
the transmitted amplitude is too high, cross-talk can be introduced onto other lines, if too low it will
not be reliably detected. If it is not possible to measure the received amplitude then it is best to
start by setting the value to l hl bo0.
10.2.10 Specific E1 configuration
10.2.10.1 E1 Line matching
For the Vega E1 product the receiver sensitivity needs to be configured based on the line length
between the Vega and the closest repeater or other ISDN endpoint.
The configuration is achieved using:
[e1t1/bri.port.n]
e1_r x_shor t _haul =0 or 1 ; 0=l ong haul and 1=shor t haul
Long haul should be selected when the cable between the Vega and the closest repeater or other
ISDN endpoint introduces more that 6dB attenuation.
Short haul should be selected when the cable between the Vega and the closest repeater or other
ISDN endpoint introduces less than or equal 6dB attenuation.
10.3 ISDN Specific Configuration
10.3.1 Introduction
ISDN signalling is a CCS (Common Channel Signalling) scheme, which means that it uses
messages in the D channel to signal call states. With a message based structure, many useful
indicators can be passed, including information like DDI, DNIS, Answer and Disconnect.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 200 - 6/2/2009
10.3.2 ISDN Network Type, Topology and Line Encoding
The following table can be used as a guide when setting up parameters for ISDN installations:
Product Physical
Connection
Topology Network E1T1/
BRIs
Line Encoding Framing Calls
Vega 400
E1:
E1-2.044 Mbps E1 Euro ISDN 4 HDB3 CRC4 /
PCM30
8 to 120
T1: T1-1.544 Mbps T1 NI2, AT&T
5ESS, DMS,
DMS_M1
4 B8ZS / AMI SF(=D4) /
ESF
8 to 92
Vega 50 Europa S/T 384 Kbps S Euro ISDN 2, 4 or
8
AZI - 4, 8 or 16
10.3.2.1 DMS-Meridian-specific ISDN setting (SIP builds only)
The e1t 1/ br i . net wor k configuration parameter has been extended to include dms_m1. This is
the selection required when connecting a SIP Vega 400 to a Meridian PABX.
The protocol implemented for this selection is identical to DMS100 (net wor k=dms) with the one
exception:
The final Channel Number Octet of the Channel ID Information Element is set to a 0 and not
1.
10.3.3 NT/TE Configuration
Each ISDN physical interface or E1T1/BRI (digital subscriber line) can be software configured to be
either the TE (Terminal Equipment) or NT (Network Termination) end. This enables the Vega to be
used in multiple scenarios, i.e. trunks plugged into a CO (Vega trunks configured as TE), trunks
plugged into a PBX (the Vega acting as though it were a CO - Vega trunks configured as NT), or
with one trunk plugged into the CO and one into a PBX. The latter scenario allows the Vega to be
inserted into an existing telephony link between a CO and PBX and based on dial plan rules, it can
either continue to pass calls between the PBX and the CO, or groom off some of the calls and
route them on as VoIP calls.
When configuring TE and NT, the value of the cl ock_mast er parameter should also be checked.
Usually, if NT is set, then cl ock_mast er should also be set, and if NT is clear (TE mode) then the
Vega should be a clock slave (cl ock_mast er =0).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 201 - 6/2/2009
The pinouts for TE and NT connections are different. On the Vega 400 the hardware pinouts
change as TE or NT are selected. In this case a straight cable can in general be used to connect
to the far end device.
10.3.4 Specific BRI configuration
1. Do not be surprised if, even after configuration, the LCD call
count remains at - - and the Trunk LED flashes indicating no
layer 2 connection. Many BRI connections do not bring up
layer 2 until a call is made.
2. Vega 50 BRI units all have 100 ohm termination impedances
across their LINKSs. Ideally the Vega should be connected
physically at the end of the LINKS.
10.3.4.1 BRI Point-to-Point Mode
Basic Rate ISDN lines (S0 bus interfaces) can be configured in one of two ways, either
Point-to-Point or Point-to-Multipoint.
Point-to-Point (PP) is used
i when a Vega is connected to a BRI CO network line which is configured to support just
one device connected directly to it (the Vega will be configured as TE) e.g. ISDN data
line connection.
ii when a Vega is the only device connected directly to a BRI PBX and is acting like a CO
network (the Vega will be configured as NT).
Point-to-Multipoint (PMP) is used
i when a Vega is connected as the NT device connected to one or more ISDN telephones
or other TE endpoints.
ii when a Vega is connected as an attached device to an S0 bus interface on a PBX or BRI
CO network where ISDN telephones would normally be plugged
Devices that are connected together on a single BRI S0 bus
must either:
- all be configured as Point-to-Point or must
- all be configured as Point-to-Multipoint.
The default mode of operation for the BRI product is to use Point-to-Multipoint mode (PMP) on all
ports.
Each PORT of the Vega 50 BRI can be independently configured to use either Point-to-Point mode
(PP) or Point-to-Multipoint mode (PMP) whether the PORT is configured as TE or NT.
In PP mode a maximum of one device at a time can be connected to each PORT. A fixed
Terminal Endpoint Identifier (TEI) must be defined for the Vega PORT, and this must match the
one configured in the corresponding device (typically configure TEI=0). Either the same or
different TEIs may be defined for each physical PORT.
The configuration parameters to set up a fixed TEI to xx on PORT n are as follows:
[bri.port.n]
l i ne_t ype=pp
t ei =xx
NOTE
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 202 - 6/2/2009
To revert the BRI back to Point-to-Multipoint mode (PMP) configure the parameter as follows:
[bri.port.n]
l i ne_t ype=pmp
(In pmp mode the value of t ei is ignored.)
10.3.4.2 BRI TE Telephone number to accept
In a Point-to-Multipoint configuration the NT device may be connected to multiple TE devices.
When a call arrives the NT device broadcasts the details of the call (including the called number) to
the TE devices. Any TE device that is configured to accept calls for that number will start ringing.
When a TE device answers the call, it locks out the other TE devices from this call and a 1:1
connection is made between the NT and the answered TE for the rest of the call.
If a Vega is one of the TE endpoints, then the parameter that configures which called number(s) it
will respond to is:
[bri.port.x.group.y]
dn
If the value of dn matches the last digits of the called number then the Vega will try to handle the
call (it will use its dial plan to onward route the call).
By default dn=*, and so the Vega will respond to every call that is sent from the NT.
Example:
If . 1. gr oup. 1. dn=34 then the Vega will respond to calls on BRI 1 to:
01344 784 934, and
020 1234 34, etc.
but will not respond to:
01344 784 933, or
020 1234 35.
dn may take the value of *, or may be a sequence of digits.
10.3.4.3 BRI Layer 2 handling
In most signalling scenarios it is required that signalling layers come up in order and that if a layer
fails, all layers are cleared down before being restarted. With certain BRI system implementations
however, the network is configured to drop L2 when not in use (but not layer 1) layer 2 is then
re-established when a call is to be made. In this case it is valid to allow layer 2 to be
re-established without layer 1 going down then up.
Vega 50 BRI units may be configured to only start layer 2 after layer 1 has just come up, or allow
layer 2 to be re-established at any time after a layer 2 disconnect. The parameter is:
[_advanced.isdn]
r est ar t _l 2_af t er _di sc=1 / 0
If set to 1 (default) the Vega 50 BRI allows re-establishment of layer 2 after a layer 2 disconnect
has occurred.
If set to zero then establishment of layer 2 is only attempted if layer 1 has just come up.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 203 - 6/2/2009
10.3.5 Verifying ISDN IEs (Information Elements)
The ISDN stack in the Vega verifies that IEs found in the signalling match the relevant signalling
specification. It verifies both the IE types, and also their content.
Where the signalling does not completely adhere to the appropriate specification the Vega can be
configured to disable this checking:
set _advanced. i sdn. ver i f y_I Es=0 disables checking of IE types (and contents
of those IEs)
set _advanced. i sdn. ver i f y_I E_cont ent s=0 disables checking of contents of IEs
See also section 10.5 Tunnelling signalling data for details on passing extra signalling information
through the Vega.
10.3.6 Call Hold
When configured as NT, BRI gateways will respond to received ISDN HOLD or SUSPEND
messages and will place the other call party on hold. The call will be taken off hold on reception of
a RETRIEVE or RESUME message. Whilst the call is on hold the tone defined by
t ones. suspended_seq will be played to the on-hold party.
10.4 QSIG Specific Configuration
10.4.1 Introduction
QSIG is a CCS (Common Channel Signalling) protocol similar to ISDN, though more tailored to
PBX to PBX communications, supporting supplementary services that enable PBXs to pass
information between themselves. Many of the same features and parameters used in configuring
ISDN signalling are also used for configuring QSIG.
QSIG is supported on E1/T1 equipped Vegas, SIP Vegas support QSIG Basic Call handling; H.323
Vegas support both QSIG Basic Call handling and QSIG tunnelling.
By enabling QSIG Basic Call handling, this allows the Vega to operate at the Q-reference point to
any Basic Call compliant device (PINX). In this mode the Vega can only send and receive the
subset of Q.931 call control messages defined in the QSIG Basic Call Specification (ISO/IEC
11572).
From details on H.323, QSIG tunnelling, see 10.5 Tunnel l i ng si gnal l i ng dat a
QSI G Tunnel i ng.
10.4.2 QSIG Network Type, Topology and Line Encoding
The following table can be used as a guide when setting up parameters for QSIG installations:
Product Physical
Connection
Topology Network E1T1
s
Line Encoding Framing Calls
Vega 400-PRI
E1:
E1-2.044 Mbps E1 QSIG 4 HDB3 CRC4 /
PCM30
8 to
120
T1 T1-1.544 Mbps T1 QSIG 4 B8ZS/AMI SF/ESF 8 to 92
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 204 - 6/2/2009
10.4.2.1 E1 QSIG Operation
The following parameters are used to configure the interface:
[e1t1]
t opol ogy=E1
net wor k=qsi g
l i ne_encodi ng=hdb3
f r ami ng=cr c4/ pcm30
[_advanced.isdn]
qsi g_mode=cont i guous/ non_cont i guous
10.4.2.1.1 E1 QSIG, Contiguous / Non-Contiguous Channel Mapping
QSIG User Channels (Uqs) can be numbered in two ways:
i) in a contiguous block, Uqs =1..30 (Uq channels 1-15 map on to Timeslots 1..15,
and Uq channels 16..30 map onto Timeslots 17-31).
ii) In a non-contiguous block, Uqs =1..15 and 17..31 (Uq channels 1-15 map directly
on to Timeslots 1..15, and Uq channels 17..31 map directly onto Timeslots 17-31).
The numbering scheme (qsi g_mode) configured on the Vega must match the scheme used by
the QSIG device that the Vega is connected to.
10.4.2.2 T1 QSIG Operation
The following parameters are used to configure the interface:
[e1t1]
t opol ogy=T1
net wor k=qsi g
l i ne_encodi ng=b8zs/ ami
f r ami ng=esf / sf
10.4.2.2.1 T1 QSIG, Contiguous / Non-Contiguous Channel Mapping
Unlike E1, there is no similar concept of contiguous / non-contiguous mapping of QSIG user
channels (Uqs).
For T1 Uqs always form a contiguous block, which maps directly onto the timeslots. (Uq channels
1..23 map onto Timeslots 1..23).
10.4.3 NT/TE or Master/Slave Configuration
Each E1T1 (digital subscriber line) can be software configured to be either QSIG master (A-side)
or QSIG slave (B-side). The nt configuration parameter is used to select the appropriate setting.
The Vega E1T1 should always be configured to be the opposite value to that configured on the
attached QSIG device. (i.e. if attached QSIG device is Master, Vega must be set to slave).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 205 - 6/2/2009
[e1t1.port.n]
nt =1 ; QSI G, mast er or A si de
[e1t1.port.n]
nt =0 ; QSI G, sl ave or B si de
NOTE
In Vega statistics A-side is indicated as NT and B-side is indicated
as TE.
When configuring A-side and B-side, the value of the cl ock_mast er parameter should also be
checked.
On the Vega 400 the hardware pinouts change as TE or NT are selected. In this case a straight
cable in general can be used to connect to the far end device.
10.4.4 Overlap Dialling
See paragraph in Error! Reference source not found. Error! Reference source not
found..
10.4.5 Type of Number configuration
Type of Number is configured as described in section 8.11 National / International Dialling Type
Of Number, but as the configuration was implemented for ISDN rather than QSIG, ISDN names
need to be used when configuring QSIG PNP TON values. When configured for QSIG signalling
the following mapping occurs:
Required QSIG PNP TON Binary Code Configuration value needed
(ISDN TON)
Unknown 0 0 0 Unknown
Level 2 Regi onal Number 0 0 1 I nt er nat i onal Number
Level 1 Regi onal Number 0 1 0 Nat i onal Number
PI SN speci f i c number 0 1 1 Net wor k- speci f i c number
Level 0 Regi onal Number 1 0 0 Subscr i ber Number
10.4.6 Message Waiting Indication
The Vega can now pass MWI (message waiting indication) as follows:
SIP to QSIG (i.e. from a SIP IP voicemail system to legacy PBX)
QSIG to SIP (i.e. from legacy PBX to SIP)
Both support for standard and Ericsson proprietary message format has been added.
The following parameters are relevant for message waiting delivery:
Parameter:
_advanced. i sdn. mwi . t ype
Possible values:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 206 - 6/2/2009
nor mal Def aul t - Use st andar d message f or mat f or MWI
er i csson Use Er i csson pr opr i et ar y message f or mat
Parameter:
_advanced. i sdn. mwi . er i csson. ASF_I E_I D
Possible values:
Def aul t 127 - Any val ue bet ween 0 and 255
Parameter:
_advanced. i sdn. mwi . er i csson. PBX_Pr ot ocol _I D
Possible values:
Def aul t 254 - Any val ue bet ween 0 and 255
Parameter:
_advanced. i sdn. mwi . er i csson. syst em_I D
Possible values:
Def aul t 0 - Any val ue bet ween 0 and 99
10.4.7 QSIG Un-Tromboning
Un-Tromboning, also known as TBCT (Two Bearer Channel Call Transfer), or call optimisation is
now supported on Vega 400s running SIP firmware. Where a call has been established through
the Vega then subsequently transferred or forwarded, the situation can exist where a trombone (or
hairpin) exists between the Vega and PBX such that two bearer channels are taken up by a single
call.
The following scenarios are supported:
Vega initiated un-tromboning, see diagram below. Un-tromboning initiated by the Vega on QSIG
so that the call is directly connected by the PBX.
PBX initiated un-tromboning, see diagram below. Un-tromboning initiated by the PBX, resulting
in the transmission of SIP REFER message so that two IP endpoints are directly connected..
Both support for standard and Ericsson proprietary message format are supported.
Vega Initiated Un-Tromboning
pbx
SIP
IP Phone
A
B
C
KEY
A calls B
B transfers to C
B hangs up A and C talk
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 207 - 6/2/2009
The Vegas default behaviour relies on detecting that two SIP call legs have the same call ID in
order to initiate the QSIG side un-tromboning. Other headers can be checked and verified using
the following parameters:
Parameter:
_advanced. si p. l oopback_det ect i on. si p_header
Possible Values:
St r i ng up t o 31 char act er s Def aul t NULL. Thi s i s t he header t o
l ook f or t o check f or a SI P l oopback cal l
Parameter:
_advanced. si p. l oopback_det ect i on. si p_header _r egex
Possible Values:
St r i ng up t o 127 char act er s Def aul t NULL. For mat i s i n t he f or m
of a r egul ar expr essi on - t he user must use < and > del i mi t er s t o
f i nd t he uni que component wi t hi n t he SI P header .
The flexible approach of specifying a regular expression was chosen as it allows other loopbacks
to be detected when interacting with other third party devices.
Example Using non-Call ID Detection
In this example the following SIP header is sent to the Vega:
TWI D: TW- CALL- SERVER- 00000108- 48d11387: - T2
Its this header rather than the Call ID header which needs to be used to detect a SIP loopback. In
this case the Call ID is different for the two legs of the call (so cannot be used for detection).
To detect the TWID header the following settings would be used:
set . _advanced. si p. l oopback_det ect i on. si p_header =TWI D
set . _advanced. si p. l oopback_det ect i on. si p_header _r egex=<TW\ - CALL\ -
SERVER\ - . *>: . *
In this case the Vega will look for two calls where the TWID header has the same content.
Everything from the start of the TWID header up to (but not including) the :-T2. The position of
the < and > indicate the section the vega will use for comparision.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 208 - 6/2/2009
PBX Initiated Un-Tromboning
10.4.7.1 Configuration
The following parameters control Un-Tromboning:
Parameter
e1t 1. por t . 1. i sdn. unt r omboni ng_enabl e
Possible values:
0 Def aul t Do not al l ow un- t r omboni ng
1 Enabl e un- t r omboni ng
Parameter:
_advanced. i sdn. unt r omboni ng. t ype
Possible values:
st andar d Def aul t Use st andar d message f or mat f or un- t r omboni ng
er i csson Use Er i csson pr opr i et ar y message f or mat
Parameter:
_advanced. si p. l oopback_det ect i on
Possible values:
0 Def aul t Di sabl e SI P l oop det ect i on
1 Enabl e l oop det ect i on f or SI P cal l s
10.5 Tunnelling signalling data
10.5.1 QSIG Tunneling (H323 Only)
QSIG is often used to connect PBXs together where advanced features, like camp-on-busy on
another PBX are required. Traditionally leased TDM lines (T1 or E1) would be used to directly
connect each PBX to each and every other PBX (a fully meshed network).
As TDM leased lines are expensive people are looking to use VoIP instead.
pbx
SIP
QSIG
IP Phone
C
KEY
A calls B
B transfers to C
B hangs up A and C talk
IP Phone
A
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 209 - 6/2/2009
QSIG tunnelling is a special mode of Vega operation whereby instead of interpreting each
signalling setup and cleardown message and converting it to an H.323 call setup or cleardown, the
Vega tunnels all D-channel (signalling) messages to their appropriate destinations. This means
that not only call setups and cleardowns can be passed across the VoIP link, but so can other
messages, such as those that allow un-tromboning of calls, those that allow camp-on-busy and
those that allow the message-waiting-indicator to be illuminated on a phone attached to a different
PBX. In this way all inter PBX communication functionality is preserved, whereas in standard
H.323 VoIP the advanced features would be lost.
Another major benefit of the VegaStream implementation of QSIG tunnelling (that follows ECMA
333) is that instead of requiring 1 E1 or 1 T1 trunk between each and every other PBX in the
network, the meshing can be carried out on a per channel basis across the IP network. Each PBX
has one Vega (or more dependent only on the simultaneous call requirement) attached to their
QSIG interface(s). For each and every signalling message the Vega will route the message to the
appropriate destination.
QSIG tunnelling is configured on a per trunk (e1t1) basis; to enable QSIG tunnelling, firstly
configure the trunk for QSIG signalling, then set the following parameter to on_demand:
[e1t1.port.n.group.m]
t unnel _mode=on_demand ; set i t t o of f t o di sabl e t unnel i ng.
For QSIG tunnelling, the dial plan needs to be configured to route calls from the telephony
interface(s) to the appropriate IP address of the far end gateway any of the usual Tokens, like
TEL: can be used in the srce statement to select the appropriate destination IP address.
NOTE
In QSIG tunnelling mode, because the QSIG signalling messages
are tunnelled through the Vegas (and not translated to H.323), the
dial plans are just used to select the destination interface and
where appropriate the destination IP address. Trying to change
for instance the TEL: or TELC: in the dial plan will not work in
QSIG tunnelling mode because the Vega does not change the
content of the messages.
For calls from the LAN interface, the dial planner just needs to select the appropriate QSIG trunk to
which to route the call.
NOTE
With the VegaStream implementation, as well as tunnelling QSIG
messages, in on_demand tunnelling mode the Vega will tunnel
any Q.931 messages.
See table in section 10.5.3 Tunnelling full signalling messages and IEs in ISDN (ETSI, ATT, DMS,
DMS-M1, NI, VN 3/4) and QSIG for details of interactions of various parameters with
tunnel_mode.
10.5.2 Tunnelling Non-QSIG Signaling Messages (H323 Only)
As QSIG is a relatively modern signaling scheme, although some manufacturers claim their PBX to
PBX protocol to be QSIG, and although most of it is, some inter-PBX messages remain
proprietary. Vegas can be configured to support this too, but because of their proprietary nature,
the Vega cannot decode each and every proprietary message. The Vega is therefore limited to
tunneling these proprietary messages on a point to point basis.
Proprietary messages still support a standard header which identifies the protocol being used in
the message. The Vega looks at the protocol ID and uses this to decide how to route the message
Vegas can route different protocols to different destinations.
The routing is carried out by the dial planner, but the details to present to the dial planner are
configured in a set of parameters as follows:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 210 - 6/2/2009
[_advanced.dsl.port.X.tunnel_protocol.Y]
cpn=of f / cal l ed_par t y_number _st r i ng
where X is the DSL port on which the proprietary message is arriving and Y is the protocol ID+1
(plus 1 so that protocol ID 0 can be handled)
When a message arrives the Vega looks at the protocol ID. If it is 8 (Q.931) then it will tunnel it
fully this is QSIG/Q.931. If it is other than ID 8, then it will use the ID+1 to index into
[_advanced.dsl.port.X.tunnel_protocol.Y]
If there is no entry, or cpn=of f , then the message will be discarded.
If cpn=cal l ed_par t y_number _st r i ng then this cal l ed_par t y_number _st r i ng will be
presented to the dial planner to obtain the routing information (IP address of the destination). The
cal l ed_par t y_number _st r i ng can consist of TEL: and TELC: tokens.
WARNING!
Where call SETUP messages are in proprietary messages, the Vega
cannot decode them, and so does not know to open a B channel (media
channel), so although the messaging may work no audio connection will be
made.
For this reason, do not include Y =9 (Protocol ID=8 Q.931 / QSIG) in the
set of [_advanced.dsl.port.X.tunnel_protocol.Y] as this will
make the Vega treat this as a proprietary protocol and so it will not
interpret the SETUP message and so will not open a media channel when
required.
Protocol Ids and Y values:
Protocol ID Y Comments
0 1 User-specific protocol
1 2 OSI high layer protocols
2 3 X.244
3 4 Reserved for system management convergence function
4 5 IA5 characters
5 6 X.208 and X.209 coded user information
7 8 Rec. V.120 rate adaption
8 9 Q.931/I.451 user-network call control messages
16 thru 63 Reserved for other network layer or layer 3 protocols, including
Recommendation X.25
64 thru 79 National use
80 thru 254 Reserved for other network layer or layer 3 protocols, including
Recommendation X.25
Other values Reserved
See table in section 10.5.3 Tunnelling full signalling messages and IEs in ISDN (ETSI, ATT, DMS,
DMS-M1, NI, VN 3/4) and QSIG for details of interactions of various parameters with
tunnel_mode.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 211 - 6/2/2009
10.5.3 Tunnelling full signalling messages and IEs in ISDN (ETSI, ATT, DMS, DMS-M1, NI, VN 3/4)
and QSIG
When passing calls from ISDN to ISDN, ISDN to / from H.323 and ISDN to / from SIP, by default
Vega gateways tokenise certain IEs (Information Elements) from the incoming signalling mesages
and re-generate the outgoing messages from those tokens. This allows the dial planner and other
Vega configuration parameters to modify the values, e.g. Calling Party Number, Called Party
Number, Display, and Bearer Capability.
Where signaling messages or specific IEs need to be passed through, the Vega can be configured
to accommodate this. This table applies to PRI and BRI signaling schemes.
e1t 1/ br i . por t . x. gr
oup. y. t unnel _mode
e1t 1/ br i . por t . x. gr
oup. y. t unnel _I Es_o
nl y
_advanced. i sdn.
I Es_t o_t unnel
Action
Off - - No tunnelling
0 - ISDN to ISDN full
message
tunnelling is not
supported
ISDN
to
ISDN
on_demand
1
N.B. Enable this parameter
on both source AND
destination trunks
Comma separated
list of IEs to tunnel
Tunnel listed IEs
Off - - No tunnelling
0 - ISDN / QSIG
tunnelled over
H.323
ISDN
to
H.323
and
H.323
to
ISDN
on_demand
1 - ISDN tunneling of
IEs not supported
over H.323
off No tunnelling
0 - ISDN tunneling
over SIP not
supported
ISDN
to
SIP
and
SIP
to
ISDN
on_demand
1
N.B. Enable this parameter
on both source AND
destination gateways
Comma separated
list of IEs to tunnel
Tunnel listed IEs
Example IE ids:
08 =cause
1c =facility
1e =progress indicator
20 =network specific facilities
24 =terminal capabilities
28 =display
29 =date and time
2c =keypad facility
34 =signal
40 =information rate
6d =calling party subaddress
71 =called party subaddress
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 212 - 6/2/2009
78 =transit network selection
7c =Low Layer Compatibility
7d =High Layer Compatibility
7e =User to User Information
96 =shift
See section 10.3.5 Verifying ISDN IEs (Information Elements) for details on how to stop the Vega
complaining about unusual Information Elements in messages.
The IEs can be tunnelled across SIP either using X-UUI headers or using a special content type
appl i cat i on/ vnd. ci r pack. i sdn- ext . This is selectable using the
_advanced. si p. q931. t x_t un_mode parameter.
Setting _advanced. si p. q931. t x_t un_mode to r eg_ur i uses X-UUI headers in SIP
messages to transport the tunnelled IEs. The preferred solution is to set
_advanced. si p. q931. t x_t un_mode to ci r pack, which causes the Vega to pass data using a
content type: appl i cat i on/ vnd. ci r pack. i sdn- ext .
10.6 CAS T1 Specific Configuration
T1 Vegas support T1 CAS (Robbed Bit Signalling) operation. In this mode each T1 trunk supports
up to 24 simultaneous calls. The specific varieties of CAS RBS supported are:
E&M Wink Start
E&M Wink Start with feature group D
FXS Loop Start
FXS Ground Start
The variety of CAS signalling to be used can be specified on a per-dsl basis. In band DTMF or MF
tone signalling is used to pass dialling information such as B-party number (DNIS), and where
supported A-party number (ANI).
10.6.1 RBS CAS Network Type, Topology, Signal type and Line Encoding
The following table can be used as a guide when setting up parameters for QSIG installations:
Product Physical
Connection
Topology Network Signal E1T1
s
Line Encoding Framing Calls
Vega 400-T1 T1-1.544 Mbps T1 RBS em_wink,
loopstart,
gndstart, fgd
4 B8ZS/AMI SF/ESF 8 to 96
10.6.1.1 RBS CAS Operation
The following parameters need to be configured for CAS operation
[e1t1]
net wor k=r bs ; selects CAS RBS operation
f r ami ng=aut o ; or esf or sf
l i ne_encodi ng=aut o ; or b8zs or ami
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 213 - 6/2/2009
[e1t1.port.n.cas]
si gnal =em_wi nk ; or loopstart, gndstart, or fgd (em_wink with
; feature group D)
di al _f or mat =. ; see configuring dial_format below for details
r x_di al _f or mat =. ; see configuring dial_format below for details
t x_di al _f or mat =. ; see configuring dial_format below for details
di gi t _di al _t i meout =6 ; Time after last dialled digit is received that DNIS / ANI
; are treated as complete 1-1000 seconds
i nf o=dt mf ; DTMF or MF
t one_del ay=50 ; delay after ack wink that first tone is sent, 1-1000 ms
[e1t1.port.1.group.m]
f i r st _chan=1
l ast _chan=aut o ; Check that this is auto or 24
[e1t1.port.2.group.m]
f i r st _chan=1
l ast _chan=aut o ; Check that this is auto or 24
NOTE
1. Some CAS schemes (e.g. E&M wink start) do not have a
called party alerting message call progress tones
(ringing, busy etc.) are passed in the media channel. For
the calling party to hear these, a media path must be
established well before the connect is received i.e. early
media must be supported and used on the VoIP side, e.g.
for the Vega either configure:
a) early H.245, or
b) fast start with accept_fast_start=3
2. For ground start and loop start signalling the Vega only
supports the TE/Slave side of the signalling protocol.
10.6.2 Configuring dial_format
ANI and DNIS are presented as in-band tones (DTMF or MF tones), separated by specifed
delimiter tones. The e1t 1. por t . x. cas. di al _f or mat parameter, now superceeded by
e1t 1. por t . x. cas. r x_di al _f or mat (for incoming calls) and
e1t 1. por t . x. cas. t x_di al _f or mat (for outgoing calls) allows the format of the reception and
presentation of the ANI and DNIS to be specified.
o =ANI (Callers telephone number)
n =DNIS (Called party number / Dialled number)
DTMF can use the separator characters: 0- 9, A- D, *, #, ~
MF can use the separator characters: 0- 9, K, S, ~
where ~indicates no character expected, K =MF KP tone, and S =MF ST tone.
e.g. *o#*n# indicates the sequence *, ANI digits, #, *, DNIS digits, #
By default
[e1t1.port.x.cas]
di al _f or mat =.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 214 - 6/2/2009
r x_di al _f or mat =.
t x_di al _f or mat =.
this configures the vega to automatically select an entry from the following table based on its
signalling configuration:
DTMF MF
E&M wink,
groundstart,
loopstart
*n# KnS
Fgd (e&M
winkstart with
feature group D)
*o#*n# KoSKnS
NOTE
The durations of the DTMF and MF signalling tones (and
inter-tone silence)is specified by dt mf _cadence_on_t i me
and dt mf _cadence_of f _t i me. You may wish to reduce
the default values of these parameters to around 70ms to
100ms each to speed up the signalling interchange.
10.6.3 NT/TE Configuration
E&M signalling, including feature group D is a symmetric signalling scheme, so there is no need for
NT/TE configuration. With loopstart and ground start signalling, which are non-symmetric, the
Vega only supports the TE side of the signalling, so again, the NT/TE is not configurable.
The value of the cl ock_mast er parameter does still need to be set up.and should be configured
as 1 if the device to which the vega is attached in not sourcing the clock, and should be set to 0 if
the other end is supplying the clock.
For Vega 400 the pinout is changed internally depending on the Nt/TE setting, so in general a
straight through cable can be used to connec to the far end device..
Further details of the Vega and cable pinouts may be found in the Product Details section of
the www.VegaAssist.com web site.
10.7 CAS E1 Specific Configuration
10.7.1 E1 CAS R2MFC
The only form of CAS signalling that the Vega gateways support is R2 MFC, a compelled tone
based CAS signalling.
Details on how to configure the Vega for R2MFC signalling may be found in the Information
Note Configuring R2MFC available from the www.VegaAssist.com web site.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 215 - 6/2/2009
11 POTS CONFIGURATION
Unlike digital systems which can be configured as either NeTwork side or Terminal Equipment side
through software configuration, the hardware required to implement analogue interfaces is different
depending on whether the gateway is to connect to telephones or whether the gateway is to
connect as though it were a set of telephones. The two types of analogue interface are known as
FXS (Subsciber / Phone facing like lines from the PSTN or extension port interfaces on a PBX)
and FXO (Office / Network facing like a bank of telephones).
Therefore, with analogue gateways the type and number of analogue ports must be specified when
ordering the product as it is not user configurable.
11.1 FXS Supplementary Services
A number of supplementary services are supported, these are:
Call Transfer
Three Way Call (3 Party Conference)
Call Forward
Do Not Disturb (DND)
Call Waiting
11.1.1 Call Transfer
See IN27 FXS Call Transfer, available on www.vegaassist.com for details on this feature.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 216 - 6/2/2009
11.1.2 Three Way Calling
When two calls are connected to an FXS port it is now possible to configure the gateway to allow
for the three calls to be connected (conference call). This feature is only available for SIP firmware
builds.
Depending on the configuration, the three-way call can be initiated by the FXS user using
command mode (a sequence of digits are dialled to initiate the three-way connection) or using
simple mode (a number of hook-flashes are performed to initiate the three-way connection).
The three way call can be initiated using two different call flow scenarios:
Call Transfer
Call Waiting
Sample Network Diagram
LAN
FXS
SIP PHONE SIP PHONE
A C
B
11.1.2.1 Command Mode / Call Transfer Three Way Call
A (SIP Phone) calls B (Analogue Phone connected to FXS port)
A connects to B
B performs a hookflash, dials C (SIP Phone)
B connects to C
B can perform further hookflashes to toggle between A and C
B enters command mode string (by default this is *54)
A, B & C enter Three Way Call
11.1.2.2 Command Mode / Call Waiting Three Way Call
A (SIP Phone) calls B (Analogue Phone connected to FXS port)
A connects to B
C calls B (B hears Call Waiting beep)
B performs a hookflash and connects to C
B can perform further hookflashes to toggle between A and C
B enters command mode string (by default this is *54)
A, B & C enter Three Way Call
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 217 - 6/2/2009
11.1.2.3 Simple Mode / Call Transfer Three Way Call
A (SIP Phone) calls B (Analogue Phone connected to FXS port)
A connects to B
B performs a hookflash, dials C (SIP Phone)
B connects to C
B performs a further hookflash
A, B & C enter Three Way Call
In Simple Mode the following number of hookflashes result in the following call connections:
First hookflash =talk to 1st caller
Second hookflash =talk to 2nd caller
Third hookflash =conference
Fourth hookflash as first hookflash
11.1.2.4 Simple Mode / Call Waiting Three Way Call
Call Waiting Three Way Call initiation is not supported when the Conference mode is Simple.
11.1.2.5 Three Way Call Indications
When switching to talk to the 1
st
caller the FXS user should hear a single short beep just before
being connected.
When switching to talk to the 2
nd
caller the FXS user should hear two short beeps just before
being connected.
When switching to talk in conference mode the FXS user should hear a single long beep just
before being connected.
11.1.2.6 Configuration
All Supplementary Service configuration can be performed via the Web User Interface (WUI). The
following parameters are accessible via the Command Line Interface (CLI).
Overall activation of Supplementary Services is enabled using the following parameter:
suppser v. enabl e
Where the parameter value can be :
0 =Disable supplementary services.
1 =Enable supplementary services (default setting).
The call conference mode is defined by the following parameter:
suppser v. pr of i l e. 1. cal l _conf er ence_mode
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 218 - 6/2/2009
Where the parameter value can be:
cmd_mode =Use command mode (dialled digit command) to initiate conference call.
si mpl e =Use simple mode (hookflashes) to initiate conference call.
The call conference command is defined by the following parameter:
suppser v. pr of i l e. 1. code_cal l _conf er ence
Where the parameter value can be:
A st r i ng of bet ween 1 and 9 char act er s (these characters must be diallable digits).
The default string is *54
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 219 - 6/2/2009
11.1.3 Call Forwarding
Call forwarding can optionally be enabled for FXS ports. Three variants are available:
Call Forward No-Answer (CFNA)
Call Forward Busy (CFB)
Call Forward Unconditional (CFU)
Call forwarding can be programmed using the handset or via CLI commands. Optionally call
forwarding statuses can be saved and restored to a server.
When a call is forwarded the dial plans are used in order to try to route the call.
When call forwarding is enabled, when going off-hook, the POTS user will hear 3 short dial tone
bursts, followed by a short pause, followed by the normal dial tone (or stutter dial tone).
11.1.3.1 Operation Examples
(Assuming default configuration, as below)
To set Call Forward Always with destination 555:
1. lift handset on POTS port
2. dial *72 555 #
This means that all calls for POTS port 1 will get forwarded to tel number 555.
To disable Call Forward Always:
1. lift handset on POTS port
2. dial *73
To enable Call Forward Always without altering call forward destination
1. lift handset on POTS port
2. dial *72 #
N.B. Call forward destinations are the same for all call forwarding.
i.e. you can't have different call forward destinations for different types of call forwarding.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 220 - 6/2/2009
11.1.3.2 Parameters
Configuring DTMF codes for call forward enable / disable:
suppser v. pr of i l e. 1. code_cf b_on Def aul t *90
suppser v. pr of i l e. 1. code_cf b_of f Def aul t *91
suppser v. pr of i l e. 1. code_cf na_on Def aul t *92
suppser v. pr of i l e. 1. code_cf na_of f Def aul t *93
suppser v. pr of i l e. 1. code_cf u_on Def aul t *72
suppser v. pr of i l e. 1. code_cf u_of f Def aul t *73
suppser v. pr of i l e. 1. code_di sabl e_al l Def aul t *00
(for all of these, default is as above but will allow any 9 character string)
11.1.3.3 Enabling call forward:
Parameter added:
pot s. por t . x. cal l _f wd_enabl e
Possible values:
on Def aul t Al l ow speci f i ed por t t o act i vat e cal l f wd
of f Do not al l ow cal l f or war d on speci f i ed por t
Parameter added:
_advanced. pot s. save_pot s_user _st at us
Possible values:
of f - Def aul t Do not save st at us t o ser ver
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 221 - 6/2/2009
f t p Save st at us t o FTP ser ver
11.1.3.4 CLI Commands - Call Forwarding Control
f xs cf dest - USAGE: f xs cf dest <por t > <cal l f wd dest or NULL>
f xs cf u - USAGE: f xs cf u <por t > <on/ of f >
f xs cf b - USAGE: f xs cf b <por t > <on/ of f >
f xs cf na - USAGE: f xs cf na <por t > <on/ of f >
Examples:
admi n >f xs cf dest 1 555
port 1, set call forward destination to 555
admi n >f xs cf u 1 on
port 1, enabled call forward unconditional
admi n >f xs cf u 1 of f
port 1, disabled call forward unconditional
admi n >f xs cf dest 1 NULL
port 1, clear call forward destination
11.1.3.5 CLI Commands - Call Forward Status Using " show ports
To query call forward status:
admi n >show por t s
Physi cal por t s:
Name Type St at us
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
POTS- 1 POTS ( FXS) on- hook r eady ( cf u, dest =555)
POTS- 2 POTS ( FXS) on- hook r eady
This shows that a call forward unconditional has been set with destination 555.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 222 - 6/2/2009
11.1.3.6 Call Forward Status - Preservation After Reboot
Config Variable:
_advanced. pot s. save_pot s_user _st at us=of f or f t p
def aul t i s " of f "
If set to "ftp", then "call forward" and "do not disturb" status will be attempted to be stored to the
configured FTP server.
Then on a reboot, the file will be read from the FTP server.
The filename will take the format XXXXXXXXXXXXfxsstatY.txt
where:
XXXXXXXXXXXX is the 12 character serial number of the unit
Y is a number representing the FXS port number
For example: 005058020604fxsstat2.txt
_advanced. pot s. save_pot s_user _st at us=of f or f t p
def aul t i s " of f "
If set to "ftp", then "call forward" and "do not disturb" status will be attempted to be stored to the
configured FTP server.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 223 - 6/2/2009
11.1.4 Do Not Disturb
Do Not Disturb (DND) can optionally be enabled for FXS ports. DND can be programmed using
the handset or via CLI commands. Optionally call forwarding statuses can be saved and restored
to a server.
When call forwarding is enabled, when going off-hook, the POTS user will hear 3 short dial tone
bursts, followed by a short pause, followed by the normal dial tone (or stutter dial tone).
The Vega can be configured to either send a busy message or ringing indication back to the calling
party.
11.1.4.1 Operation Examples
(Assuming default configuration, as below)
To set Call Forward Always with destination 555:
1. lift handset on POTS port
2. dial *78
This means that all calls for POTS port 1 will get forwarded to tel number 555.
To disable Call Forward Always:
1. lift handset on POTS port
2. dial *79
11.1.4.2 Configuration Parameters
suppser v. pr of i l e. x. code_dnd_on Def aul t *78
suppser v. pr of i l e. x. code_dnd_of f Def aul t *79
(for all of these, default is as above but will allow any 9 character string)
Parameter added:
pot s. por t . x. dnd_enabl e
Possible Values:
on Def aul t Al l ow DND t o be act i vat ed f or speci f i ed por t
of f DND cannot be act i vat ed f or speci f i ed por t
Parameter added:
pot s. por t . x. dnd_of f _hook_deact i vat e
Possible Values:
on Goi ng of f - hook i mmedi at el y cancel s DND
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 224 - 6/2/2009
of f Def aul t Goi ng of f - hook does not cancel DND
Parameter added:
pot s. por t . x. dnd_r esponse
Possible Values:
i nst ant _r ej ect Def aul t - I nst ant _r ej ect cal l wi t h SI P 480 or SI P
message as def i ned by do_not _di st ur b. st at us_code
spoof _r i ngi ng Send r i ngi ng back t o cal l or i gi nat or
Parameter added:
_advanced. si p. do_not _di st ur b. st at us_code
Possible Values:
400- 699 - Def aul t 480 SI P st at us code t o use f or DND
Parameter added:
_advanced. si p. do_not _di st ur b. st at us_t ext
Possible Values:
St r i ng up t o 47 char act er s, def aul t " Do Not Di st ur b"
11.1.4.3 CLI Commands - DND Control
f xs dnd - USAGE: f xs dnd <por t > <on/ of f >
Example:
admi n >f xs dnd 1
port 1, enabled do not disturb
11.1.4.4 CLI Commands - DND Status Using " show ports
To query DND status:
admi n >show por t s
Physi cal por t s:
Name Type St at us
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
POTS- 1 POTS ( FXS) on- hook r eady ( dnd)
If DND has been activated, the "(dnd)" text will be present
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 225 - 6/2/2009
11.1.4.5 DND Status - Preservation After Reboot
See entry under Call Forward for details.
11.1.5 Call Waiting
When a call is placed into an FXS port that already has an active call the Vega (if configured) plays
a call waiting indication tone to the FXS port and sends a SIP 180 or 183 message to the new
caller to indicate ringing. Optionally the Vega can now be configured to send a SIP 182 Queued
message so that the caller is aware of the status of the call.
Parameter:
_advanced. si p. cal l _wai t i ng. st at us_code
Possible values:
of f Def aul t Use SI P 180 / 183 as nor mal
182 Use SI P 182 Queued f or cal l wai t i ng cal l
See IN38 FXS Call Waiting for more information on this feature.
11.2 POTS Phone Facing (FXS) ports
FXS ports on a Vega gateways are designed to connect to conventional, loop start POTS
telephony products such as telephones and faxes; also to connect to analogue trunk interfaces of
PBXs. Operation of the interface involves the following activities:
11.2.1 DTMF digit detection
DTMF Digits are detected automatically by the Vega and no parameters are necessary to
configure this operation.
11.2.2 Hook Flash detection
The maximum period of time for detecting a line break as a hookflash (as opposed to on-hook) is
configured in
[_advanced.pots.fxs.x]
hookf l ash_t i me
Typically, values of between 100ms and 800ms are appropriate.
If the call clears when hookflash is being detected, then increase the value of hookf l ash_t i me.
Also see:
[_advanced.pots.fxs.x]
hookf l ash_debounce_t i me
11.2.3 Ring Cadence Generation
Each POTS port can generate a number of different (or distinctive) outgoing ring patterns. A
different ring pattern can be referenced (r i ng_i ndex) for each different group section created
for the FXS POTS port concerned. The ring cadence generator uses the r i ng_i ndex to select a
particular ring pattern as defined in _advanced. pot s. r i ng. x.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 226 - 6/2/2009
E.g. The following parameters would be used to configure the Vega such that whenever an
outgoing call is presented to FXS interface 33 the ring pattern is defined by the first entry in the ring
cadence table:
[pots.port.n.if.m]
r i ng_i ndex=1
i nt er f ace=33
[_advanced.pots.ring.1]
f r equency=50
name=I nt er nal - UK
etc.
11.2.4 Line supervision Answer and disconnect
Loop Current disconnect
FXS ports on Vega gateways can be configured to provide a Loop Current Disconnect signal on
their FXS ports when calls cleardown on the LAN side. To configure Loop Current Disconnect
generation on FXS ports, use the following parameters:
[_advanced.pots.fxs.1]
l oop_cur r ent _br eak
l oop_cur r ent _del ay
l oop_cur r ent _t i me
l oop_cur r ent _br eak is the overall enable / disable flag, l oop_cur r ent _t i me is the time that
the loop current will be broken for (make sure that this is slightly longer than the attached devices
detection period). l oop_cur r ent _del ay is a configurable delay after the other party has
cleared that the Vega waits before issuing the loop current disconnect; this gves the FXS party a
chance to clear the call before the loop current disconnect is issued.
NOTE
Whilst the loop current disconnect is being issued, there is no line
voltage / current to detect, and so no other POTS events can be
detected, for example, on-hook and off-hook events can not be
detected until completion of the loop current disconnect.
Line Current Reversal
FXS ports may be configured to reverse the line voltage on the POTS interface on call answer and
call disconnect. To enable this function set:
[_advanced.pots.fxs.x]
l i ne_r ever sal =1
WARNING!
If the Vega is configured to operate using line current reversal
then the device which is attached to the Vega must also
support this functionality as answer and cleardown are
indicated using the line current reversals.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 227 - 6/2/2009
11.2.5 DTMF digits after answer
Vega FXS ports can be configured to send DTMF digits after answer in order to further route the
call on the connected device.
This feature is controlled by the TEL: token in the destination dial plan entry; if a dial plan entry that
routes calls to an FXS port has a TEL: token containing some digits, when the FXS port is taken
off-hook the DTMF will be played out.
e.g.. if the following dial plan routes the call:
sr ce=I F: 99. . , TEL: <501> dest =I F: 0101, TEL: <1>
the Vega will play out the digits 501 immediately after the call is answered on port FXS 1.
11.3 POTS Network Facing (FXO) ports
FXO ports on a Vega gateways are designed to connect to an analogue CO switch or analogue
extension ports on a PBX.
11.3.1 Line voltage detection
Before an outbound call is made Vega FXO ports check that there is line voltage on the line. If no
line voltage is observed (less than +/- 5volts) the call is rejected with cause code 27; this can be
checked for in the dial planner / call presentation group and used to represent the call to another
destination which is active.
11.3.2 Impedance configuration
The impedance of the FXO ports is configurable from the user interface (both web browser and
CLI). Three choices of impedance are selectable:
1. 600R (US style)
2. CTR21 (European style)
3. 900R
NOTE
Although in practice the Vega will operate when the impedance is
set incorrectly, for approvals reasons it is important that you
configure the FXO port to the impedance utilised by the country in
which the Vega is installed. For example:
600R CTR21
Canada, Caribbean,
Central America, China, Hong
Kong, Malaysia, Mexico,
Saudi Arabia, South America,
Taiwan, Thailand, United Arab
Emirates, United States
Austria, Belgium, Cyprus,
Denmark, Finland, France,
Germany, Greece, Iceland,
Ireland, Israel, Italy,
Liechtenstein, Luxembourg,
Netherlands, Norway,
Portugal, Spain, Sweden,
Switzerland, United Kingdom
FXO port impedance is configured in the FXO Port Hardware Configuration Profile parameters:
[_advanced.pots.fxo.y]
i mpedance
On the web browser, change it in the FXO Parameters section of the POTS >Advanced POTS >
FXO Configuration >Hardware Profile Configuration (Modify)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 228 - 6/2/2009
Ensure that the hardware profile associated with the port has the correct impedance set. The
hardware profile selection for each FXO port is made in:
[pots.port.x]
f x_pr of i l e
Set f x_pr of i l e=y
On the web browser, this is found it in the Modify Port section of the POTS >Port Configuration
(Modify)
11.3.3 DTMF digit generation
The DTMF on/off times, initial holdoff between off-hook and dialling, and DTMF tone amplitude are
all user configurable:
[_advanced.pots.fxo.x]
dt mf _hol dof f _t i me=200
[_advanced.dsp]
dt mf _gai n=10000 - bei ng super ceeded by dt mf hi / l o gai n
dt mf _hi _gai n
dt mf _l o_gai n
dt mf _cadence_on_t i me=150
dt mf _cadence_of f _t i me=250
It is strongly recommended that the values of dt mf _hi _gai n and dt mf _l o_gai n are not
altered; changing these value from default may cause the Vega to produce out-of-spec DTMF
tones
11.3.4 Hook Flash generation
The time period for generating the hookflash (on-hook) pulse is configured in
[_advanced.pots.fxo.x]
hookf l ash_t i me
Typically a value of around 500ms is appropriate.
11.3.5 Ring Cadence Detection
FXO ports on a Vega gateway are only capable of detecting a single incoming ring pattern. The
following parameters are used to configure the cadence detection circuit for a particular ring:
[_advanced.pots.fxo.x]
r i ng_det ect _l ongest _r i ng_of f =5000
r i ng_det ect _shor t est _r i ng_on=250
Examples:
Parameter UK USA
Longest silence 2000ms 4000ms
Shortest ring 400ms 2000ms
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 229 - 6/2/2009
11.3.6 Line Supervision Answer and Disconnect
Vega FXO ports operate in one of three modes for line supervision.
1) No Supervision
Disconnect Supervision: In this mode the Vega FXO port is unaware of the on-hook/off-
hook state of the far end during a call. The responsibility for tearing down a call lies with
the VoIP side of the call, regardless of which end established the call. Usually the VoIP
subscriber will hear the other party hang up followed by call progress tones indicating that
the far end caller has disconnected; they will then hang up the call in response.
Answer Supervision: When an outgoing call is attempted over the FXO interface the
Vega will connect and answer the incoming VoIP call at the same time as dialling out on
the POTS line. If billing is carried out based on the VoIP messaging, callers will be charged
for outdialling and any following success or failure messages there is no answer signal
available to be passed through the Vega.
2) Loop Current Detection:
Disconnect Supervision: In this mode the Vega FXO port detects the short break in loop
current which the PBX / CO switch generates (to indicate that the far end party has
terminated the call) and it will clear the call through itself.
Answer Supervision: This method does not indicate that the far end has answered the
call. When an outgoing call is attempted over an FXO interface the Vega will connect and
answer the incoming VoIP call at the same time as dialing out on the POTS line. If billing
is carried out based on the VoIP messaging, callers will be charged for outdialling and any
following success or failure messages there is no answer signal available to be passed
through the Vega.
Loop Current disconnect detection is enabled by setting:
[_advanced.pots.fxo.x]
l oop_cur r ent _det ect =loop_current_disconnect_time
The loop_current_disconnect_time value should be configured to be slightly
shorter than the period for which the PBX / switch makes the break in loop current.
NOTE
The l oop_cur r ent _det ect time MUST be greater
than hook_f l ash_t i me, otherwise a hook flash will
cause the call to clear down.
3) Line Reversal Detection:
Disconnect and Answer Supervision: In this mode the FXO port detects the polarity of
the line to determine if the far end has answered the call and also uses it to sense if the far
end has terminated the call. When an outgoing call is attempted over the FXO interface
the Vega will only connect the incoming VoIP side if the far end answers (indicated by the
line current being reversed to its active state).
Call cleardown is indicated by the line current being reversed back to its idle state. If line
reversal is supported by the CO Switch/PBX then it allows the Vega to answer the call
when the destination call is answered and the Vega to clear the call when the destination
call is cleared. If billing is being carried out on the VoIP messaging then the caller will
correctly only be billed for the voice connected part of the call.
It is enabled by setting:
[_advanced.pots.fxo.x]
l i ne_r ever sal _det ect =1
Other parameters associated with line current reversal are:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 230 - 6/2/2009
[_advanced.pots.fxo.x]
l i ne_r ever sal _sampl e_del ay=<t i me>
l i ne_r ever sal _debounce_t i me=<debounce t i me>
WARNING!
If line_reversal is enabled on a Vega FXO port but is not
supported by the PBX / switch that it is connected to, then
outgoing FXO calls will never be answered (as there will never
be a line reversal)
If possible either loop current detection or line reversal should be used to ensure calls are cleared
from FXO ports in a timely manner. However only one method of supervision should be enabled at
a time enabling them both is likely to stop the Vega handling calls correctly.
11.3.7 Tone Detection
The method of tone detection configuration described in this section is available on Vega 50
Europas. Vega 5000s use a different method, described in IN36 Configuring Ans_n_Disc
Supervision available at www.vegaassist.com.
If no other means of reliable disconnection signalling are available (i.e. battery line
reversal or loop current disconnection signalling) and progress tones are provided
(i.e. busy, congestion and disconnection indications) a Vega gateway can be
configured to detect disconnection tones which are received on an FXO port.
It is useful to think of an FXO interface / port as an analogue handset when
considering call supervision.
For an inbound call, as ringing voltage is received into an FXO interface, the port will
go 'off-hook'. Depending on the dial plan configuration the inbound call maybe routed
immediately to a destination interface or secondary dial tone may be played to the
calling party (who is making the calling 'into' the FXO port).
For an outbound call, as a call is routed (via the dial plan) to the FXO interface, the
port goes 'off-hook' and plays DTMF tones to the exchange / pbx (i.e. the called
number is dialled). At this point of the call the calling leg of the call will
automatically be connected, i.e. if the calling party is SIP a 200 OK is sent
immediately to the calling party.
Once the inbound (or outbound) call is terminated by the PSTN / PBX party (or the
call fails to establish as the destination is busy or congested), disconnection tones are
played towards the FXO interface. If configured to do so, the FXO interface will
detect these tones and the FXO port will go 'on-hook' ready for another call.
11.3.7.1 Configuration
Firstly, if tone detection is going to be used as the method for call disconnection
ensure that all other disconnection methods are disabled. The following parameters
values disable all other disconnection methods:
_advanced. pot s. f xo. 1. l i ne_r ever sal _det ect =0
_advanced. pot s. f xo. 1. l oop_cur r ent _det ect =0
_advanced. pot s. f xo. 1. voi ce_det ect =0
The following parameters determine the FXO interface tone disconnection
configuration (and activation):
_advanced. pot s. f xo. x. t onedet ect
Where 'x' represents the FXO profile in use by a specific port. Possible values are:
0 (default) - disconnection tone detection is disabled.
1 - disconnection tone detection is enabled.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 231 - 6/2/2009
Busy, Congestion and Disconnection tones can (each optionally) be detected by
configuring the following parameter set values:
t onedet ect . x. y. enabl e
t onedet ect . x. y. f r eq1
t onedet ect . x. y. f r eq2
t onedet ect . x. y. f r eq3
t onedet ect . x. y. of f _t i me1
t onedet ect . x. y. of f _t i me2
t onedet ect . x. y. of f _t i me3
t onedet ect . x. y. on_t i me1
t onedet ect . x. y. on_t i me2
t onedet ect . x. y. on_t i me3
Where:
x =busy, congestion or disconnect
y =profile index - i.e. if two different busy tones need to be detected a profile can be
created for each type of tone detection, i.e. tonedetect.busy.1 and tonedetect.busy.2 etc.
In the majority of cases only one profile needs to be configured for each
disconnection tone type (busy, congestion, disconnection).
t onedet ect . x. y. enabl e
Possible values are 0 or 1 - i.e. disable or enable the detection of the tone defined in
this tone detection profile.
t onedet ect . x. y. f r eq1
t onedet ect . x. y. f r eq2
t onedet ect . x. y. f r eq3
Possible values are 250 - 700, which represents a frequency (in Hz) present in the tone
defined in this tone detection profile. If the tone is single frequency the values of
freq2 and freq3 should be set to 0 - i.e. no detection.
t onedet ect . x. y. of f _t i me1
t onedet ect . x. y. of f _t i me2
t onedet ect . x. y. of f _t i me3
Possible values are 0 to 10,000, which represents the off time (in Miliseconds) of the
cadence of the tone to be detected. Tones which contain mutliple cadences can be
detected by configuring differing off_time values (i.e. off_time2 and off_time3).
Unless the tone does contain multiple cadences off_time2 and off_time3 should be set
to 0 - i.e. no multi-cadence detection.
t onedet ect . x. y. on_t i me1
t onedet ect . x. y. on_t i me2
t onedet ect . x. y. on_t i me3
Possible values are 100 to 10,000, which represents the on time (in Miliseconds) of
the cadence of the tone to be detected. Tones which contain mutliple cadences can be
detected by configuring differing on_time values (i.e. on_time2 and on_time3).
Unless the tone does contain multiple cadences off_time2 and off_time3 should be set
to 0 - i.e. no multi-cadence detection.
11.3.7.2 Detecting Tones
There are a number of commands that can be used to display the tones that are received at the
Vega FXO port. The output of these commands can be used to correctly configure the parameters
described above.
To display the frequencies and cadences that are being received the following commands can be
issued:
debug on
debug t one enabl e
When the debug t one enabl e command is issued the Vega is no longer able to detect tones
and thus disconnect calls. i.e. Its not possible to both debug and detect tones.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 232 - 6/2/2009
See below for sample output from the above commands.
To stop the debug output:
debug t one di sabl e
To query the status of the commands:
debug t one st at us
Sample Output
_DSP : Tr ace : 0145465: 07315: DSP : user - def i ned t one det ect ed di gi t 412Hz, 0Hz ( di gi t 3)
: ( dspac. c; 1036)
_DSP : Tr ace : 0145835: 00370: DSP : user - def i ned t one i s now of f : ( dspac. c; 914)
_DSP : Tr ace : 0146215: 00380: DSP : user - def i ned t one det ect ed di gi t 412Hz, 0Hz ( di gi t 3)
: ( dspac. c; 1036)
_DSP : Tr ace : 0146595: 00380: DSP : user - def i ned t one i s now of f : ( dspac. c; 914)
_DSP : Tr ace : 0146970: 00375: DSP : user - def i ned t one det ect ed di gi t 412Hz, 0Hz ( di gi t 3)
: ( dspac. c; 1036)
_DSP : Tr ace : 0147350: 00380: DSP : user - def i ned t one i s now of f : ( dspac. c; 914)
_DSP : Tr ace : 0147730: 00380: DSP : user - def i ned t one det ect ed di gi t 412Hz, 0Hz ( di gi t 3)
: ( dspac. c; 1036)
_DSP : Tr ace : 0148110: 00380: DSP : user - def i ned t one i s now of f : ( dspac. c; 914)
From the sample output above it can be seen that the detected frequency was 412Hz and the
cadence is 370ms on-time (145835 145465) and 380ms off-time(146215 145835)
11.3.8 FXO Slow network cleardown
In certain networks, for instanceMobile networks it takes a long time for the Network to clear. If a
new call is made immediately after a previous one clears, the call will fail. In order to
accommodate this, the Vega can be configured to prevent new calls to FXO ports until a specified
period has passed since the previous call cleared. To configure this, use parameters:
[_advanced.pots.fxo.x]
por t _not r el eased_cause
por t _r el ease_del ay
If a call is attempted within the port_release_delay period after the previous call cleared, then the
Vega will reject the call with cause code port_notreleased_cause. This can be used to try and re-
present the call using call re-presentation.
11.3.9 FXO Secondary dial tone
Usually an FXO interface will immediately route a call as soon as it detects ring tone.
If the dial plan specifies a TEL: token in the dial plan for an FXO port, when a call arrives at that
port, rather than routing the call immediately, dial tone will be played to the caller. The caller can
then enter digits using DTMF tones (phone key presses), and the digits received will provide digits
for the TEL: token comparison in the dial planner. Calls can now be routed using TEL:, as well as
TELC:, IF: etc.
The time that dial tone is played for (and before the call is routed assuming NO digits are entered)
is defined by:
[pots.profile.2]
dt mf _di al _t i meout =5
(this is the inter digit DTMF timeout). If the timeout is set to 0 then the call will be routed
immediately (effectively turning off the secondary dial tone feature).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 233 - 6/2/2009
11.4 Analogue Caller-ID (CLID)
Analogue Vega gateways support caller ID by receiving / generating FSK or DTMF tones during
the ringing phase of a call.
Vega FXS ports generate the tones towards the attached telephones, and FXO ports detect the
tones when they are sent by the attached PBX / CO switch.
Several types of CLID encoding are supported on the Vega units; the appropriate mechanism can
be configured by setting the parameter:
[pots]
cal l er i d_t ype=gr 30- sdmf / gr 30- mdmf / bt / et si - f sk / et si - f sk- l r
/ et si - f sk- post / et si - dt mf / et si - dt mf - l r / et si - dt mf - post / of f
gr30-sdmf
Conforms to Bellcore standard GR30 - single data message format. J ust passes the call time and
number information. The latest standard mentions that this format may be dropped in future.
gr30-mdmf
Conforms to Bellcore standard GR30 - multiple data message format. This passes the caller name
as well as the call time and number. (This configuration will also receive gr30-sdmf caller Ids)
bt
Based on the gr30-mdmf format but with a difference in the tones and interface to the POTS as
required for use in the UK. The specification requires the phone to send a whetting pulse after the
first tones are detected.
etsi-fsk
Use ETSI FSK, delivered before ring.
etsi-fsk-lr
Use ETSI FSK, delivered before ring but after line reverse.
etsi-fsk-post
Use ETSI FSK, delivered between 1
st
and 2
nd
ring.
etsi-dtmf
Use DTMF, delivered before ring.
etsi-dtmf-lr
Use DTMF, delivered before ring but after line reverse.
etsi-dtmf-post
Use DTMF, delivered between 1
st
and 2
nd
ring.
off
Turns off Caller ID handling.
The parameter
[pots.port.n]
cal l er i d
controls Caller ID on a port by port basis; it can take the values off or on.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 234 - 6/2/2009
11.4.1 FXS Outbound Analogue Caller ID (CLID) H.323 and SIP
Caller ID generation can be enabled and disabled on a per port basis using
[pots.port.n]
cal l er i d=on/ of f .
The particular line encoding type used must be set up in:
[pots]
cal l er i d_t ype= cal l er I d t ype
Caller ID is sent out both at the start of a call and, if the call waiting supplementary service is
enabled, when a 2
nd
call arrives mid call
11.4.2 FXO Analogue Caller ID detection (CLID) H.323 and SIP
Incoming caller id is configured using 3 parameters,
[pots.port.n]
cal l er i d = on/ of f
[pots]
cal l er i d_t ype = cal l er i d t ype
cal l er i d_wai t = t i me t o wai t t o see i f a cal l er I D i s bei ng
pr esent ed i f t i me i s exceeded t hen t he Vega
assumes t hat no cal l er I D wi l l be r ecei ved.
Vega FXO ports do not support the generation of caller ID.
Some caller ID generation methods provide no warning that caller ID is about to be delivered. i.e.
there is no initial ring splash or line whetting pulse. For these installations the Vega can now
allocate a DSP resource to permanently listen for caller ID tones.
The Vega will only allocate a permanent DSP resource where there is line voltage present on the
FXO port (i.e. there is a connected device) and the configured caller ID type doesnt provide any
warning of caller ID delivery. i.e. One of the following types of caller ID is configured:
et si - f sk
et si - dt mf
WARNING!
This permanent allocation may affect the ability of other
ports on the gateway to complete calls. This affects
gateways where there are both FXS and FXO ports.
11.4.2.1 SIP Presentation Field
This presentation field address extension may be present in the From: header of an INVITE as:
presentation = ( anonymous | public | unavailable)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 235 - 6/2/2009
If caller ID is on, the caller ID will be displayed (passed on) if:
There is NO presentation address extension in the From: header of the INVITE message
The INVITE messages presentation is public
Caller ID WILL NOT BE DISPLAYED (will not be passed on) if:
The INVITE messages presentation is unavailable, in which case the phone will display
OUT OF AREA
The INVITE messages presentation is anonymous, in which case the phone will display
BLOCKED CALL
If there is no caller ID to put in the From: field (none supplied, presentation restricted etc.) then
Unknown will be used.
See also RPID handling in section Error! Reference source not found. Error! Reference
source not found.
11.4.2.2 H.323 extensions
Additional parameters are available to configure the text of the messages that are sent over H.323
under specific received caller ID situations:
[advanced.h323control]
nocal l er i d=<no caller id text>
not avai l =<no caller id available text>
r est r i ct ed=<caller id is restricted text>
11.5 Power fail fallback operation
Vega FXS gateways which include 2 FXO ports support power fail fallback. If the Vega is powered
down, rebooted, or in the middle of an upgrade, it will use fall back relays to connect the first two
FXS ports to the two FXO ports. This provides emergency telephony, even under VoIP-down
conditions.
On returning to an active state, the Vega samples the condition of the FXS <-- >FXO lines, if
either are in use, it will delay removing the relay connection until both are free.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 236 - 6/2/2009
12 H.323 CONFIGURATION
H.323 variants of the Vega gateway are designed to operate in one of two modes:
Gatekeeper mode
Standalone mode (no gatekeeper)
In Gatekeeper mode, at power up or re-boot the Vega will register with the gatekeeper, and then
for each call the Vega will send the call details (like called number, calling number, name and if
appropriate TA: and TAC:) to the gatekeeper and the gatekeeper will carry out the authentication,
routing and translation, providing the Vega with destination dialled number, name and if
appropriate TA: information.
In standalone mode, the Vega dial planner effectively implements a subset of gatekeeper
functionality, carrying out the authentication, routing and translation internally.
Therefore, when a gatekeeper is used, the dial planner is typically much simpler than for
standalone mode as the gatekeeper will do the number translations etc.
T
h
e
r
e
a
r
e
s
e
v
e
r
a
l
F
i
r
s
To select the mode of operation configure h323. gat ekeeper . enabl e on the CLI or select the
appropriate Gatekeeper Mode or Standalone button on the H.323 page on the web browser
interface.
Dial
Planner
Gatekeeper
Whitelist
Standalone
mode
Incoming
Telephony Call
Outbound
VoIP Call
Whitelist
Dial
Planner
Gatekeeper
Standalone
mode
Incoming VoIP
Call
Outbound
Telephony
Call
Incoming VoIP
Call
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 237 - 6/2/2009
12.1 Standalone Mode
In standalone mode (h323. gat ekeeper . enabl e=0) the Vega dial planner needs to be
configured to contain all operations for authentication, routing and translation.
Details on configuring the dial planner can be found in section 8 The Dial Planner.
In some cases it is required that most calls are to be routed to the same destination on the LAN
(e.g. another gateway); to do this, a default H.323 endpoint address can be set up. This endpoint
address is used in all cases where an explicit ongoing IP address is not specified in the dial plan
entry.
[h323.if.x]
def aul t _i p=www. xxx. yyy. zzz
def aul t _por t =1720
NOTE
For readability, it is recommended that the TA: token is used
explicitly in all dial plan entries rather than using the
def aul t _i p parameter
12.2 Gatekeeper Mode
In gatekeeper mode (h323. gat ekeeper . enabl e=1) a number of parameters need to be set up
to allow registration and authentication to take place with the gatekeeper. Specifying which
gatekeeper to use is carried out by either specifying a static IP address/host name, or by enabling
auto-discovery. In the latter case a multicast is used to find the nearest gatekeeper.
To enable auto-discovery set:
[h323.gatekeeper]
aut o_di scover =1
For manual discovery a gatekeeper IP address needs to be specified:
[h323.gatekeeper]
aut o_di scover =0
def aul t _gat ekeeper =www. xxx. yyy. zzz
In either case, during the registration process a number of identifiers (alias) may be sent from the
Vega to the gatekeeper to allow authentication of the Vega and to identify which calls the Vega can
handle. Each alias can be an email address, a URL, an H.323 id or an E.164 number
For example:
[h323.gatekeeper.terminal_alias.n]
t ype=h323
name=Vega
Check with your system administrator to see what authentication aliases are required by the
gatekeeper. Most gatekeepers require either an H.323 ID or a list of E.164 prefixes.
NOTE
1. Setting h.323.gatekeeper.terminal_alias_n.name to
NULL means do not send this terminal alias.
2. Terminal aliases are re-registered with the
gatekeeper on APPLYing changes
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 238 - 6/2/2009
Some gatekeepers decide which calls to route to a gateway based upon the telephone number
prefixes that the gateway can handle. In the gatekeeper registration process the Vega will declare
all the telephone number prefixes defined in dial plan entries for srce expressions for the LAN
interface (IF:05). A telephone number prefix is the fixed length expression before a .* in a TEL:
token.
e.g. 01344 will be declared as a prefix for the dial plan entry:
sr ce=I F: 0501, TEL: 01344. *
NOTE
1. Dial plan prefixes are re-registered with the
gatekeeper on APPLYing changes
2. For Cisco call manager prefixes need to be
preceded by a #. In the Vega dial planner duplicate
each prefix dial plan entry and put a #after the
TEL: (before the dialled number prefix).
12.3 Gatekeeper Registration Status Command and Messages
To monitor the progress of the Vegas registration with the Gatekeeper a number of LOG
messages are logged. They are of the form:
LOG: 03/ 04/ 2001 14: 06: 42 H323 ( A) Rb6C00 GK st at e xxx ( event yyy)
The gatekeeper state values can be:
I dl e ; gat ekeeper i s not r egi st er ed
Di scover ed ; gat ekeeper i s trying t o r egi st er
Regi st er ed ; gat ekeeper i s r egi st er ed
If the Vega is configured to be in gatekeeper mode it will only make (or receive) VoIP calls when
the gatekeeper status is Registered. To obtain the current registration status, use the CLI
command:
gat ekeeper st at us
12.4 Gatekeeper Registration Commands
A number of CLI commands are available to request the Vega to un-register / register with the
gatekeeper.
gat ekeeper unr egi st er
- forces the gateway to unregister with the gatekeeper
gat ekeeper r egi st er
- forces the gateway to send a registration request to the gatekeeper
gat ekeeper r er egi st er
- forces the gateway to unregister from the gatekeeper and then register with the gatekeeper.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 239 - 6/2/2009
12.5 Fast Start
Fast start (or fast connect) is a feature of H.323 which simplifies and speeds up the connect
procedure by reducing the number of messages exchanged between the endpoints on making a
call. Fast start was added to the H.323 standard at version 2.0 and is not compatible with the
earlier version 1.0 H.323 standard. For this reason it is not supported by all H.323 endpoints (and
so this feature may sometimes need to be turned off on the Vega).
By default a Vega will accept all incoming fast start connections and will attempt to initiate fast start
for outgoing H.323 calls.
The operation of fast start on the Vega can be controlled using the following parameters:
[h323.profile.x]
use_f ast _st ar t =1
accept _f ast _st ar t =1
h245_af t er _f ast _st ar t =1
use_f ast _st ar t controls whether the Vega initiates outgoing H.323 calls requesting
fast start.
accept _f ast _st ar t controls whether the Vega will accept fast start information or whether
it will force the sender to use Version 1.0 H.323 call setup interactions.
The parameter value defines when the faststart will be accepted 3 =in
the CALL PROCEEDING message, 2 =in the ALERTING message,
1=in the CONNECT message. If, for example, the parameter is set to
3 and no call proceeding is sent, then the fast start accept will be sent
with the alerting or if there is no alerting, it will be sent with the
connect.
h245_af t er _f ast _st ar t controls whether a channel is created for media control during fast
start. Usually fast start chooses not to open a separate media
signalling channel, but with this value enabled it will do so if requested
by the other endpoint. (The H245 media control connection is required
for Out-of-band DTMF)
12.6 Early H.245
Early H.245 is a feature that allows a voice path (or media channel) to be created between two
H.323 endpoints before the call has been accepted. This has many advantages over establishing
the media channel after successfully connecting:
Call progress tones from the B-party can be heard during call setup (e.g. ringback)
Call progress tones from the B-party can be heard during unsuccessful call setup (e.g. busy
tone, recorded announcements)
Call connection times are reduced because the media channel has already been connected
before the user answers
This is a Version 2.0 H.323 feature and is therefore only compatible with other Version 2 compliant
endpoints. To control the use of early H.245, the following configuration parameters have been
provided:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 240 - 6/2/2009
[h323.profile.x]
use_ear l y_h245=0
accept _ear l y_h245=1
The default behaviour is to accept early H.245 if it is requested, but not to initiate it for outgoing
calls.
12.7 H.245 Tunnelling
H.245 tunnelling reduces the number of TCP/IP connections made per call by eliminating the need
for separate sockets for both call signalling (Q.931) and channel signalling (H.245). This feature
can be enabled and disabled for both incoming and outgoing calls independently as follows:
[h323.profile.x]
use_h245_t unnel =0/ 1 [ def aul t =1]
accept _h245_t unnel =0/ 1 [ def aul t =1]
useindicates use tunnelling for outgoing H.323 calls,
accept indicates allow tunnelling on incoming H.323 calls.
The default configuration is that this more efficient mode of operation is enabled for both outgoing
and incoming calls.
NOTE
If the called/calling H.323 endpoint does not support h.245
tunnelling then, even with use/accept enabled the call will
automatically proceed by connecting an H.245 socket as though
H.245 tunnelling were disabled.
12.8 Round trip delay
Round trip delay monitoring is used to check whether a LAN connection is lost during a VoIP
conversation. This is especially useful for wireless endpoints which may go out of wireless range
during the call if the round trip delay messaging stops getting a response, the call is cleared
down with a configurable cause code. Round trip delay is configured using the following
parameters:
[_advanced.h323]
r t d_f ai l ur e_cause=41 ; RTD f ai l ur e cause code
[h323.profile.x]
r t d_i nt er val =0 ; I nt er val bet ween sendi ng RTD
; r esponse r equest s
r t d_r et r i es ; Number of t i mes t o r et r y
; r esponse r equest bef or e
; f ai l i ng l i nk
[_advanced.rad.h245]
r oundTr i pTi meout =5 ; Ti me t o wai t l ooki ng f or RTD
; r esonse see r oundTr i pTi meout
12.8.1 Round trip delay (RTD) operation
Although round trip delay is configured on a per unit basis, round trip delay testing is carried out on
a per call basis. So, for every active call:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 241 - 6/2/2009
when round trip delay is enabled (r t d_i nt er val <>0) at every r t d_i nt er val period
an RTD request response (like a ping) is sent out to the endpoint associated with this call
the Vega waits r oundTr i pTi meout time for a reply after sending the RTD request
response; if it is not received within the specified time the Vega increments the RTD fail
count for that call, if the response is received within the r oundTr i pTi meout time, then
the RTD fail counter for that call is cleared
if the RTD fail count exceeds the retry count (r t d_r et r i es) the link is deemed to have
failed and the call is cleared down and the reason for cleardown given as
r t d_f ai l ur e_cause.
Typically, if an endpoint is going to respond to the RTD response request, it will do so promptly, so
r oundTr i pTi meout can be set smaller than r t d_i nt er val .
NOTE
In practice, if round trip delay monitoring is not enabled, or the
delays for RTD detection are long, the TCP socket will timeout
and break the signalling connection.
12.9 H.450 for Call Transfer / Divert
12.9.1 Introduction
H.450 is the set of standards used by H.323 to provide Supplementary Service Support.
H.450.1 H.450 Series Title
H.450.2 Call Transfer
H.450.3 Call Diversion
H.450.4 Call Hold
H.450.5 Call Park/Pickup
H.450.6 Call Waiting
H.450.7 Message Waiting Indication
H.450.8 Name Identification Service
H.450.9 Call Completion on Busy Subscriber
H.450.10 Call Offer
H.450.11 Call Intrusion
12.9.2 H.450.2 Call Transfer
H.450.2 provides the capability to transfer calls. It provides mechanisms for one party (the
transferring party) to instruct a remote party (the transferred party) with which it is currently in a
call, to be transfered to a third party (the transferred-to party).
If the call transfer is actioned when the transferring party is in a call with the transferred-to party,
this is known as a transfer with consultation.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 242 - 6/2/2009
If the transferring party is not already in a call with the transferred-to party then the transfer is
known as a transfer without consultation.
12.9.2.1 Transferring Party Support
Vegas do not support the functionality of a transferring party. i.e. There is no support for initiating
transfer requests.
12.9.2.2 Transferred-to party support
Incoming calls specifying that they are H.450.2 transfers will be accepted. There is however no
support for Transfer with Consultation.
12.9.2.3 Transferred party support
During an active call a transfer instruction from the remote endpoint (transferring party) will cause
the Vega to initiate a new outgoing call to the specified destination (transferred-to party).
If the transferred-to party supports H.450.2 the original call will be released when the
transferred-to party accepts the transfer. If this is before the transferred-to party call is
connected a ringback tone will be played to the transferred party.
If the transferred-to party does not support H.450.2 the original call will only be released
when the transferred-to call is connected.
Transfers with Consultation will be accepted provided that the Transferring party does not require
any specific support from the Vega gateway while it makes the consultation call.
12.9.3 H.450.3 Call Diversion (For test purposes only)
NOTE
This feature has not been fully released and therefore
should only be used in test lab environments
H.450.3 provides the capability to forward calls before they are answered. It provides a
mechanism for a called endpoint (Diverting Party) to instruct the calling endpoint (Diverted Party) to
divert the call to a third endpoint (Diverted-to Party). Reasons for diversion are controlled by the
Diverting Party and can include Divert on Busy, Divert on No Answer, Always Divert.
12.9.3.1 Diverting Party
Vegas do not support the functionality of a diverting party. i.e. There is no support for initiating
divert requests.
12.9.3.2 Diverted-to Party
The Vega will accept calls diverted-to it, however there is no support for informing the diverted-to
party that this is a diverted call or the reason for the call diversion.
12.9.3.3 Diverted Party
All diversion reasons will be accepted and a redirected call generated. Multiple redirections are
supported, ie if Vega A calls endpoint B, which redirects to C it is possible for C to re-divert to D
(resulting in a call A to D)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 243 - 6/2/2009
12.9.4 H.450 Configuration
[serviceprofile.n]
defines the Supplementary services that are to be supported. This allows up to 10 distinct profiles
to be defined. Each profile has the following parameters:
[serviceprofile.n]
name ; a text identifier
t r ansf er ; 0 =do not support call transfer, 1 =support call transfer
di ver t ; 0 =do not support call diversion, 1 =support call diversion
t r ansf er _cal l er _i d ; = t r ansf er r i ng_par t y / t r ansf er r ed_par t y
defines which caller ID is displayed when a call is
transferred to the Vega.
Changes to serviceprofile parameters take immediate effect, being used for the next call that uses
the corresponding profile.
The default configuration contains a single profile in which all services are enabled.
[h323.if.x]
serviceprofile
is an integer that selects the service profile to be used for H.323 calls. If this value is set to zero all
supplementary services are disabled for H323. Otherwise the corresponding serviceprofile defines
which supplementary services will be enabled. It is made effective using the APPLY command.
The default configuration is ser vi cepr of i l e=0, i.e. supplementary services are disabled.
[_advanced.h450]
contains some general parameters and sections for each supported standard. All parameter under
here are effective on save and reboot.
[_advanced.h450]
max_cal l s
max_ser vi ces
these parameters control the amount of resource that the Radvision stack will allocate to support
the H.450 functions.
[_advanced.h450.h450_2]
t i mer _ct - t 1=20
t i mer _ct - t 2=22
t i mer _ct - t 3=24
t i mer _ct - t 4=26
these parameters are timers for H450.2
[_advanced.h450.h450_3]
t i mer _t 1=20
t i mer _t 2=22
t i mer _t 3=24
t i mer _t 4=26
t i mer _t 5=28
these parameters are timers for H450.3
All these parameters should only be altered from their default values on advice from VegaStream
engineers.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 244 - 6/2/2009
13 MEDIA
The following codecs are supported:
G711ALaw
G711ULaw
G729
G723.1
GSM-FR
T.38
ClearMode
Both RTP (Real Time Protocol) and SRTP (Secure Real Time Protocol) are supported.
13.1 Media Channels and CODECs
13.1.1 H.323 Media Channels and CODECs
In the process of making an H.323 VoIP call, (i.e. a call to IF:0501) each endpoint sends a list of
codecs that it supports (a capability set list) to the other endpoint involved in the call. The order
in which the codecs are listed defines the desired priority of use. The first codecs are the most
preferred, and the last listed codec is the least preferred. The two endpoints then independently
choose one of the offered codecs to use to send their audio.
Depending on the type of service being provided a different set of codecs may need to be offered,
or at least the preferred priority order of the codecs may need to be altered.
The list of voice codecs that an H.323 Vega gateway offers, and the priority order in which they are
offered is affected by the version of code, the mode of operation, and a number of configuration
parameters.
Vega gateways use different parameters to select the codecs to offer depending on whether the
mode of operation is fast-start or not. For example, a small set of codecs can be offered on an
initial fast-start, with perhaps a wider range then offered if the fast-start negotiations fail.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 245 - 6/2/2009
Faststart:
Non faststart:
dial plan CAPDESC
CAPDESC=0 No CAPDESC speci f i ed CAPDESC=n
( n > 0)
H323.profile.x.faststart_capset
f ast st ar t _capset =0 f ast st ar t _capset =n
( n > 0)
media.capset.n.caps
caps=a, b, , c
Voice Codecs offered
Al l medi a. cap. x
ent r i es
Sel ect i on of Voi ce
Codecs:
medi a. cap. a,
medi a. cap. b,
,
medi a. cap. c
speci f i ed by t he
medi a. capset . n. caps
l i st
dial plan CAPDESC
CAPDESC=0 No CAPDESC speci f i ed CAPDESC=n
( n > 0)
H323.profile.x.capset
capset =0 capset =n
( n > 0)
media.capset.n.caps
caps=a, b, , c
Voice Codecs offered
Al l medi a. cap. x
ent r i es
Sel ect i on of Voi ce
Codecs:
medi a. cap. a,
medi a. cap. b,
,
medi a. cap. c
speci f i ed by t he
medi a. capset . n. caps
l i st
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 246 - 6/2/2009
In the dial planner a token CAPDESC: can be used (in a dest statement where the interface is
IF:0501) to specify which codec set (medi a. capset . n. caps list) is to be used to specify the list
of codecs to offer (and their priority order).
If CAPDESC: 0 is specified, rather than using the medi a. capset . n list, then all codecs that the
Vega has been configured to support, the whole list of medi a. cap. x entries, will be offered in the
priority order x=1 highest, x=2 second priority etc.
If the dial plan does not specify a CAPDESC: then depending on whether it is a fast-start
negotiation or not, either the parameter h323. pr of i l e. x. f ast st ar t _capset , or
h323. pr of i l e. x. capset will specify the default codec set to offer. (Note, if a a faststart
negotiation is attempted and fails causing drop-back to standard H.323 codec negotiation, or if re-
negotition of codecs is required during the call e.g. to add fax capabilities to the call then
h323. pr of i l e. 3. capset will specify the codecs offered.) If the f ast st ar t _capset , or
capset , whichever is being used is set to 0, then the selection of codecs offered will be the same
as if CAPDESC: 0 had been specified in the dial plan. If the parameter =n, where n >0 then the
selection of codecs offered will be the same as if CAPDESC: n were specified in the dial plan.
NOTE
1. Vegas do not support asymmetric codecs (i.e. different
codecs for send and receive) If this occurs with certain
endpoints, use CAPDESC to reduce the codecs offered to
those endpoints.
13.1.2 SIP Media Channels and CODECs
In the process of making a SIP VoIP call, (i.e. a call to IF:9901) the initiating end sends a list of
codecs that it supports in an SDP. (The order in which the codecs are listed defines the
preference order for usage of the codecs).
The receiving end chooses a codec that it also supports and responds with its own SDP chosing
just one of the offered codecs as the codec to use for the call.
The codecs that a Vega offers (when it sends the initial sdp) and the codecs that the Vega
compares the offered codecs list against to decide which codec to accept are configurable.
The codecs to be used are specified as follows:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 247 - 6/2/2009
In the dial planner a token CAPDESC: can be used (in a dest statement where the interface is
IF:9901) to specify which codec set (medi a. capset . n. caps list) is to be used to specify the list
of codecs to offer (and their priority order).
If CAPDESC: 0 is specified, or if the dial plan does not specify a CAPDESC: then the parameter
si p. capset will specify the codec set to offer. si p. capset can only take values >0; its value
specifies the codec set (medi a. capset . n. caps list) to be used to specify the list of codecs to
offer (and their priority order).
NOTE
1. Vegas do not support asymmetric codecs (i.e. different
codecs for send and receive).
Parameters for the individual codecs may be adjusted under the relevant sections of the DSP
configuration subsection (Media Channels section on the web browser) see section 13.3 SIP and
H.323 - Configuring CODEC Parameters.
When the SIP Vega makes a call it offers the codecs (in the same order as specified in the media
capset) to the far end gateway the far end gateway will choose one of the codecs to use. When
receiving calls, the Vega will look through the incoming list of offered codecs and will accept the
first (highest priority) offered codec which matches one of those listed in its own media capset list.
13.1.3 CAPDESC Capability descriptors list
The CAPDESC token in the dial planner provides a per-call mechanism to select the CODECs
offered over H.323 or SIP:
CAPDESC: n
dial plan CAPDESC
CAPDESC=0 or
No CAPDESC speci f i ed
CAPDESC=n
( n > 0)
sip.capset
capset =n
( n > 0)
media.capset.n.caps
caps=a, b, , c
Voice Codecs offered
Sel ect i on of Voi ce
Codecs:
medi a. cap. a,
medi a. cap. b,
,
medi a. cap. c
speci f i ed by t he
medi a. capset . n. caps
l i st
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 248 - 6/2/2009
This token, which is placed in the destination part of the dial plan entry (for calls to IF:0501 or
IF:9901), forces a particular list of CODEC types to be advertised in the capabilities for this
outgoing call. The list of the CODECs to be offered is defined in the medi a. capset . n section of
the configuration parameters, for example:
[h323.profile.x]
f ast st ar t _capset =0
capset =0
[sip]
capset =2
[media.cap.1]
codec=g7231
[media.cap.2]
codec=g711Al aw64k
[media.cap.3]
codec=g711Ul aw64k
[media.cap.4]
codec=t 38t cp
[media.cap.5]
codec=t 38udp
[media.capset.1]
caps=1, 2, 3
[media.capset.2]
caps=2, 3
In the above example the selection of medi a. capset entry 1 causes all configured codecs
(G.723.1, G.711Alaw64k and G.711Ulaw64k) to be offered. medi a. capset entry 2 however has
been restricted to offer G.711 only (A law and U law).
With this configuration, if CAPDESC: 2 is used in a dial plan destination expression it will force only
the G.711 codecs to be advertised for calls using this dial plan entry.
NOTE
The medi a. capset . n lists define both the subset of codecs to
offer and also the priority order in which they will be offered.
Vegas support both G.723.1 and G.729A (G729) compression standards at the same time, though
due to DSP memory addressing capabilities, individual DSPs cannot run code for all codecs at the
same time. The DSP memory can be loaded with code to support G.711Alaw, G.711Ulaw and
G.723.1 or G.711Alaw, G.711Ulaw and G.729A (G729).
At boot up the Vega loads different DSPs with different code images in order to reduce the
likelihood of having to load new code on the fly. The medi a. cap. n. codec entries define which
code images to load. If a codec is negotiated and there is no spare DSP resource with that code
loaded, in the background, a DSP will be loaded with the appropriate code image.
13.1.4 Defining Fax capabilities
13.1.4.1 Fax capabilities
Fax capabilities are treated as codecs. Two fax only codecs are available for H.323: t38tcp and
t38udp the TCP and UDP variants of T.38 respectively; for SIP, the specifications only define a
single codec t38udp the UDP variants of T.38.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 249 - 6/2/2009
If t38tcp and / or t38udp are to be used then medi a. cap. n entries have to be created for them.
To offer T.38 codecs for fax, add the capabilities to an appropriate media.capset.x
Whether to include the capability in the h323. pr of i l e. x. f ast st ar t _capset ,
h323. pr of i l e. x. capset , si p. capset or just in a capset that can be called up using
CAPDESC in a dial plan depends on how and when the fax codecs should be offered.
In H.323, this can depend upon the other fax devices in the network, e.g. some VoIP gateways like
to set up the fax capabilities right at the start of the call, and so in this case fax codecs should be
included in the H323 f ast st ar t _capset . Others only want to negotiate fax if and when
required; in this case do not include it in the H323 f ast st ar t _capset , but include it
in h323. pr of i l e. x. capset .
NOTE
For H323 firmware, selection of only one t.38 fax codec (either
t38udp or t38tcp) is recommended where possible many
products do not respond properly when offered more than one
fax codec, and this can lead to invalid codecs being chosen.
13.2 SIP Media Channels And CODECs
Vegas support both G.723.1 and G.729A (G729) compression standards at the same time, though
due to DSP memory addressing capabilities, individual DSPs cannot run code for all codecs at the
same time. The DSP memory can be loaded with code to support G.711Alaw, G.711Ulaw and
G.723.1 or G.711Alaw, G.711Ulaw and G.729A (G729).
At boot up the Vega loads different DSPs with different code images in order to reduce the
likelihood of having to load new code on the fly. The medi a. cap. n. codec entries define which
code images to load. If a codec is negotiated and there is no spare DSP resource with that code
loaded, in the background, a DSP will be loaded with the appropriate code image.
For details on configuring which codecs a SIP Vega will offer (and accept) when making and
receiving calls, see section 15.4.3 SIP SDP a= ptime and direction
13.3 SIP and H.323 - Configuring CODEC Parameters
Each codec has some specific parameters that can be altered. The codec parameters are
grouped under codec type. Some of them are a parameter associated with the telephony and
VoIP interfaces, others which are more call related. The two types are stored in separate areas,
dsp.xxx and media.packet.codec.y. Each parameter takes effect on the next call attempt after a
change has been made; this allows the user to tweak settings to obtain the optimal configuration
for a given situation. The available parameters are listed in the tables below.
Interface related parameters:
Parameter dsp.xxx Description Effect of increasing / enabling this
parameter
Other notes
VP_FIFO_nom_delay minimum jitter
buffer size in
milliseconds
1) improves audibility of received audio
when interworking with software based
codecs (e.g. Microsoft Netmeeting) which
introduce permanent jitter.
2) Improves audibility of received audio
when connecting over the internet, or
other data networks where there is
Set this value >=2 to 3 times
the packet time or to the
maximum observed jitter on
the LAN network plus 1
packet time (whichever is the
larger value) but do not set
it larger than it needs to be;
the larger the value the larger
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 250 - 6/2/2009
Parameter dsp.xxx Description Effect of increasing / enabling this
parameter
Other notes
significant jitter.
3) Increases the delay for the voice path
the latency, and the lower the
perceived quality.
VP_FIFO_max_delay maximum jitter
buffer size in
milliseconds
1) improves the audibility on data networks
which introduce random amounts of jitter.
2) In cases of large jitter this will increase
voice path delay
This value defines the
maximum size the FIFO can
dynamically grow to leave
this set at maximum for best
results
7
Echo_tail_size amount of
echo
cancellation
used in
milliseconds
1) eliminates echo up to length selected
2) introduces fixed length delay of length
selected
Leave at the default of 16ms
unless echo is a problem. If it
is increase to 32, 64 or 128 as
proves necessary
8
VADU_threshold silence
suppression
activation
threshold
increases the level at which the codec will
differentiate between background noise and
speech. I.e. when not to send audio and
when to send audio if VADU_enable_flag is
set.
Generally leave this as default
may need to increase if
background noise level is high
(otherwise the VAD detector
will never trigger)
idle_noise_level background
comfort noise
level
increases the level of ambient noise
generated in the listeners ear when no audio
is received from the source gateway (due to
VAD detector detecting silence and so not
sending audio packets)
Generally leave this set at the
default value
tx_gain packet transmit
gain
increases the sound level for packets
transmitted across the LAN
Typically limit gain increases
to <=7 more than this can
result in clipped audio.
On FXO units this alters the
gain of DTMF tones from the
PSTN/PBX too much
adjustment can take the tone
volumes out of spec.
rx_gain packet receive
gain
increases the sound level for packets
received from the LAN
Typically limit gain increases
to <=7 more than this can
result in clipped audio.
Per call related parameters:
Parameter
media.packet.codec.y
Description Effect of increasing / enabling this
parameter
Other notes
out_of_band_DTMF out of band
DTMF tone
enable /
disable
When enabled:
1) introduces a slight fixed delay into the
voice path
2) the Vega detects and deletes the
DTMF tones from the Audio stream that
is to be sent across the LAN it sends
messages across the signaling link to
tell the far end what DTMF tones it
detected. The far end Vega will then
re-generate the tones so that they are
pure to the destination.
Need to use
out_of_band_DTMF for
G.723.1 as it compresses
audio so much that when
audio is expanded at the far
end the tones are not
accurately reproduced.
For G.711 and G.729
out_of_band_DTMF may be
selected or not as desired.
7
If the two VoIP endpoints are not synchronised through their telecoms interfaces then slip can occur causing the fifo
buffers to run near empty then empty or near full then over full. If excessive delays are observed it may be best to reduce
the Max delay value to limit the maximum delay, BUT note that if slip occurs beyond the Max delay then audio will be lost
and intelligibility of audio will be degraded.
8
Vega 100 units require special firmware builds to support 64 and 128ms echo tail size use showdsp to see the DSP
capabilities. (Note, long echo tail size builds may limit the maximum number of simultaneous calls an E1 Vega can handle.)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 251 - 6/2/2009
Parameter
media.packet.codec.y
Description Effect of increasing / enabling this
parameter
Other notes
packet_time size of voice
packets
transmitted
by Vega in
milliseconds
1) improves reception on busy reliable
networks by decreasing the number of
packets transmitted per second
2) increases the likelihood of audible sound
loss on unreliable networks 1 packet
contains more audio
3) Reduces bandwidth required to transfer
audio
4) Increases latency
The smaller the packet time
the higher the perceived
quality due to lower latency
VADU_enable_flag silence
suppression
enable /
disable
enabling will
1) introduce a slight voice path delay
2) result in packet suppression on the
network when no-one is speaking.
Enabling this can introduce
clipping of speech if this is
observed try disabling this
feature
13.4 G.729 / G.729 Annex A/B Codecs
The G.729 Codec is variously known as G.729, G.729 Annex A and G.729 Annex B, or even
G.729 Annex A/B. G.729 is the original codec name, and also the generic name. Annex A
introduced a codec which is interoperable with G.729 but is mathematically a lot less complex
(therefore much more affordable in terms of DSP processing power). Annex B then added the
optional (programmable) silence suppression. Vega gateways use the G.729 Annex A/B version of
codec, whether the G.729 or G.729 Annex A variety is selected as it is backward compatible with
the other variants:
H.323
Two codec names G.729 and G.729 Annex A are supported by the Vega for backward
compatibility. In H.323 some products negotiate for a codec called G.729AnnexA (as
defined in the H.323 specification), others for a codec named G.729 (not per
specification). Vegas allow negotiation for both codecs. By allowing each to be selected
as a separate codec, different parameters can be provisioned for the two.
SIP
RTP/AVP in SIP sdps is configured as a numeric value, 18 for G.729. In Vega gateways
this enables a G.729 Annex A/B codec which is backward compatible with both G.729 and
G.729 Annex A. Enabling G.729 or G.729 Annex A in media.cap.n will ensure that there
are G.729 Annex A/B codecs immediately available for use (see section 13.2 SIP Media
Channels And CODECs).
NOTE
To change the parameters for the SIP G.729 codec, change the
parameters in the G.729 section (not the ones in the
G.729 Annex A section).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 252 - 6/2/2009
13.5 Out of band DTMF (OOB DTMF)
Compression CODECs such as G.723.1 and to a lesser extent G.729 distort audio because they
must lose information in order to perform the compression. For normal speech this distortion is
insignificant and hardly affects the intelligibility of the speech. However, in the case of pure tones
(such as DTMF) this distortion modifies the tones enough that they are no longer within
specification, and so DTMF detectors may not recognise the tones. The solution is to detect the
tones before the audio is compressed, remove the tones from the audio stream and send the
DTMF information as separate packets out of the audio stream to the far endpoint, which will
then generate a pure DTMF tone back into the audio stream.
Such a mechanism is known as out of band DTMF, and is supported in all Vega products (SIP and
H.323) for both transmission and reception.
By default the feature is enabled for all CODECs except G.711 A and u law (G.711 codecs will
pass DTMF tones through uncorrupted). To change the setting use the
medi a. packet . codec. y. out _of _band_DTMF parameter in the configuration database.
13.5.1 H.323 out of band DTMF
In H.323, Out-of-band DTMF information is sent in H.245 UserInputIndication messages they can
be sent in two formats: alphanumeric or simple mode, and signal mode. Vega gateways will
accept OOB DTMF messages generated in either format. By default Vega gateways will use the
signal type format to send OOB DTMF information, but this can be configured in the following
configuration parameter:
[h323.profile.x]
oob_met hod=si gnal ; al phanumer i c=al phanumer i c/ si mpl e; si gnal =si gnal ;
none=none
Alphanumeric / simple mode does not support DTMF tone duration information.
Signal mode supports optional timing information. (However, Vega gateways do not send timing
information, and ignore any received timing information).
13.5.2 SIP out of band DTMF
In SIP, Out-of-band DTMF information can either be sent in Info messages, or from using
RFC2833.
For further details on RFC 2833 see section 15.5 RFC2833
For further details on Info messages see the SIP Signalling Messages Appendix.
13.6 Tones
13.6.1 Configuring Local Call Progress Tones
During call establishment, and usually during call disconnection the caller hears call progress
tones. These tones include: busy tone, ringing tone, unobtainable, etc. Sometimes these are
generated by the Network, sometimes the Vega passes the audio through from another device and
sometimes the Vega generates the call progress tones itself.
Because each tone cadence may vary from country to country, the Vega provides a facility for the
user to change their definition. Configuration is via a three tiered set of configuration parameters,
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 253 - 6/2/2009
[tones], [tones.def] and [tones.seq]. These parameters can be configured directly through a CLI
interface or via the web browser from the menu tones.
The [ t ones] section provides a mapping of the call progress tones that the Vega offers to specific
tone sequence IDs:
[tones]
di al t one_seq=1 ; gener al di al t one f or maki ng cal l s
st ut t er d_seq=2 ; st ut t er di al t one ( not i mpl ement ed on H. 323)
busyt one_seq=3 ; busy t one on cause 17
f ast busy_seq=4 ; f ast busy t one f or number not f ound
r i ngback_seq=5 ; r i ngback t one f or f ar end r i ngi ng
cal l wai t 1_seq=6 ; cal l wai t i ng t one 1 ( not i mpl ement ed on H. 323)
cal l wai t 2_seq=7 ; cal l wai t i ng t one 2 ( not i mpl ement ed on H. 323)
The [ t ones. seq] section specifies the sequences. For each sequence ID the list of raw tones,
their duration and their order are specified. The duration value is measured in milliseconds; a
value of 0 means play the tone forever. E.g. tone sequence ID 1 plays tone 1 for 10 seconds then
tone 6 forever:
[tones.seq.1]
name=di al _seq
r epeat =0
[tones.seq.1.tone.1]
pl ay_t one=1
dur at i on=600000
[tones.seq.1.tone.2]
pl ay_t one=6
dur at i on=0
If the tones that make up the sequence are all of finite duration, the r epeat parameter defines
whether the sequence of tones are played just once in sequence (r epeat =0) or are played
repeatedly in sequence (r epeat =1).
The [ t ones. def ] section specifies the raw tones:
[tones.def.1]
name=di al t one
f r eq1=350
amp1=6000
f r eq2=440
amp2=6000
f r eq3=0
amp3=0
f r eq4=0
amp4=0
on_t i me=0
of f _t i me=0
r epeat =1
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 254 - 6/2/2009
This parameter structure allows the tone to be defined consisting of up to 4 different frequencies;
each frequency has an associated amplitude with it. Within this parameter structure it is also
possible to specify an on_t i me and an of f _t i me so that pulsed tones can be specified. If
on_t i me=0 then this means play the tone forever, if on_t i me<>0 then the of f _t i me silence
follows the on_t i me. The r epeat can be used to repeat pulsed tones.
Tone definition parameter summary:
Parameter Range Description
amp1, amp2, amp3, amp4 0-32,500 Relative amplitude
f r eq1, f r eq2, f r eq3, f r eq4 0-4,000 frequency (Hz)
Name 31 chars descriptive string
on_t i me 0-10,000 duration (ms) of tone on (0=play tone forever)
of f _t i me 0-10,000 duration (ms) of tone off
Repeat 0 (FALSE)
Or:
1 (TRUE)
for one-shot tone, set to 0.
for on_time, off_time tone cycle to repeat, set to 1.
13.6.2 Fixed Tone Table
In addition to the configurable tone table above, the Vega has a set of pre-defined tones for DTMF
and Silence. The CLI command show f i xed t ones lists the index numbers of the fixed DTMF
tones in case you ever need to use them in tone sequences.
LI ST OF FI XED TONES
- - - - - - - - - - - - - - - - - - -
name i ndex
DTMF_0 100
DTMF_1 101
DTMF_2 102
DTMF_3 103
DTMF_4 104
DTMF_5 105
DTMF_6 106
DTMF_7 107
DTMF_8 108
DTMF_9 109
DTMF_A 110
DTMF_B 111
DTMF_C 112
DTMF_D 113
DTMF_HASH 114
DTMF_STAR 115
SI LENCE 116
DTMF tones have the following characteristics:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 255 - 6/2/2009
amp1=10000, amp2=10000, on_t i me=80ms, of f _t i me=50ms, r epeat =0
13.6.3 Selecting Generation of Progress Tones vs Media Pass Through
13.6.3.1 H.323 tx_media_before_connect
The t x_medi a_bef or e_connect parameter only affects telephony to H.323 calls; it allows the
user to control whether media (RTP traffic) may be sent before answer (connect). If set to 0, then
the RTP data is not generated until a CONNECT message has been received on the H.323
interface. If set to 1, then RTP data is generated as soon as the H.323 protocol negations allow.
[h323.profile.x]
t x_medi a_bef or e_connect =0/ 1 [ def aul t =0]
NOTE
If set to 1, some software endpoints have been found to forward
the audio before the phone has been answered
13.6.3.2 SIP progress_if_media
The pr ogr ess_i f _medi a parameter allows the user to force the use of 180 Ringing (rather
than 183 Session Progress) if an ISDN ALERTING message is received with an in-band media
indicator.
It may alternatively be used to force the use of a 183 message if media is generated locally by
the Vega.
if pr ogr ess_i f _medi a=0, then 180 ringing is always used to indicate ringing (whether media
exists for the ringing cadence or not; if media exists, an sdp will be present)
if pr ogr ess_i f _medi a=1, then if media exists for the ringing a 183 Session Progress will be
used (instead of the 180 Ringing). If no media is available for ringing, (in ISDN a flag indicates
whether or not there is inband audio) then a 180 Ringing will be used. Note this acts upon the
indicator in the ISDN messaging and is not overridden by the decision to generate tones locally
(tones.net.ring=1)
if pr ogr ess_i f _medi a=2, then if media exists, either from the incoming call, or generated
locally (tones.net.ring=1) 183 with sdp will be used, otherwise if no media a 180 will be used.
In each case RTP audio will be sent as soon as SDPs are agreed and media is available.
[_advanced.sip]
pr ogr ess_i f _medi a=0/ 1/ 2 [ def aul t =2]
To see how this parameter interacts with others for an FXS interface, see table in 13.6.3.5 FXS
SIP parameters for ringback generation to the VoIP interface
To see how this parameter interacts with others for an ISDN interface see table in 13.6.3.6.1
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 256 - 6/2/2009
ISDN SIP parameters for ringback generation to the VoIP interface
13.6.3.3 Network (Remote) Call Progress Tones (SIP gateways only)
It is possible to configure a SIP Vega to generate call progress tones that are played back over the
LAN, for scenarios where it is not possible to generate the progress tones at the local end.
13.6.3.3.1 Tone Types
When configured (see section 13.6.3.3.5 Configuration Parameters for Network Tones (SIP only))
there are 3 kinds of tones that can be played:
1) ringback - normal ringback tone
2) failure - tone played when call couldnt be made e.g. due to "engaged" or "unreachable"
3) disconnect - tone played when call was hung-up at the far end first.
13.6.3.3.2 Ringback Tone
For example, when a user A makes a VoIP call to / through the Vega, he / she can hear the
ringback tone generated by the remote Vega.
User A on User B on
SI P phone- - - - - LAN- - - - - - Vega
<- - - - - - ( sends r i ngback usi ng RTP)
13.6.3.3.3 Failure Tones
For example, remote user engaged:
1) User A calls User B.
2) User B is engaged.
3) User A hears the busy tone generated by the Vega.
User A on User B on
SI P phone- - - - - LAN- - - - - - Vega
<- - - - - - ( sends busy t one usi ng RTP)
13.6.3.3.4 Disconnect Tones
For example, remote user hangs up first:
1) User A calls User B.
2) User B answers and then hangs up
3) User A hears the busy tone generated by the Vega
User A on User B on
SI P phone- - - - - LAN- - - - - - Vega
<- - - - - - ( sends busy t one usi ng RTP)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 257 - 6/2/2009
13.6.3.3.5 Configuration Parameters for Network Tones (SIP only)
Network tones are enabled using the following parameters:
[tones.net]
r i ng=1 ;set to '1' to enable playing of ringback tone towards packet network
The tones definitions used for the Network call progress tones are:
Network tone Use tone defined by
Ringback t ones. r i ngback_seq
Failure t ones. busyt one_seq
disconnect t ones. busyt one_seq
13.6.3.4 Vega FXO ringback_present
The r i ngback_pr esent parameter is designed for use on line current reversal lines to
control whether during outdial the calling party hears ringback tone, or whether they hear the
dial tone, outdial and any progress tones.
[_advanced.pots.fxo.x]
r i ngback_pr esent =0/ 1 [ def aul t =1]
If r i ngback_pr esent =0, on an FXO outbound call ringback tone is passed to the VoIP
interface until the FXO answer is received
If r i ngback_pr esent =1, on an FXO outbound call, audio from the FXO line is passed
across the VoIP interface as soon early media allows audio to be transferred
NOTE
On standard loopstart lines, the answer occurs on seizing the
FXO line, so all dialling etc. will be heard whatever the value of
this parameter. On line current reversal lines ringback tone will be
heard until answer if this parameter is set to 0.
13.6.3.5 FXS SIP parameters for ringback generation to the VoIP interface
The following table shows the interaction of various parameters with the generation of ringback
tone to the SIP interface.
Tones.net.ring _advanced.sip. progress_if_media Result
Generate ringback tone to
packet network when Alerting
0: Force use of180 if alerting
1: Use 183 rather than 180 if media present in alerting
2: Use 183 if either in-band or locally generated media
0 180 (no sdp) 0
1, 2 183 (no sdp)
0, 1 180 with sdp; Locally generated ringback 1
2 183 with sdp; Locally generated ringback
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 258 - 6/2/2009
13.6.3.6 ISDN
Configuration parameters are available to allow control over the playing of dial tone and in-band
progress tones from the Vega.
NOTE
DSLs configured as NT generate dial tone and progress tones by
default, but _advanced.isdn.force_disconnect_progress still needs
to be configured to define the maximum time to play disconnect
tone at the end of a call.
[_advanced.isdn]
user _di al t one=0/ 1 [default=0]
set to 1 configures TE E1T1s on ISDN interfaces to
originate dial tone towards an NT device.
[_advanced.isdn]
user _pr ogr ess=0/ 1 [default=0]
set to 1 configures TE E1T1s on ISDN interfaces to
originate progress tones towards an NT device, for both
DISCONNECT and ALERTING messages.
[_advanced.isdn]
al er t _wi t h_pr ogr ess=0/ 1/ 2 [default=1]
Set to 0 causes the Vega to ignore any In-band Media
indication in ISDN Alerting messages (media is not cut
through at this stage)
Set to 1 causes the Vega to act upon any In-band
Media indication in ISDN Alerting messages (media is
cut through if in-band media is indicated)
Set to 2 causes the Vega to Assume In-band Media on
receiving an ISDN Alerting message (media is cut
through immediately after the Alerting message has
been received).
[_advanced.isdn]
pr ogr ess_wi t h_pr ogr ess=0/ 1/ 2 [default=1]
Set to 0 causes the Vega to ignore any In-band Media
indication in ISDN Progress messages (media is not
cut through at this stage)
Set to 1 causes the Vega to act upon any In-band
Media indication in ISDN Progress messages (media is
cut through if in-band media is indicated)
Set to 2 causes the Vega to Assume In-band Media
on receiving an ISDN Progress message (media is cut
through immediately after the Progress message has
been received).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 259 - 6/2/2009
[_advanced.isdn]
send_pr ogr ess_as_al er t i ng=0/ 1 [default=0]
Set to 0 allows progress messages to be passed
through unchanged
Set to 1 causes received progress messages from
ISDN interfaces to be converted to alerting messages
before being forwarding onto the VoIP interface or
another ISDN interface.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 260 - 6/2/2009
13.6.3.6.1 ISDN SIP parameters for ringback generation to the VoIP interface
The following table shows the interaction of various parameters with the generation of, or
passing through of ringback tone to the SIP interface.
ISDN
messaging
_advanced.isdn.
alert_with_progress
_advanced.isdn.
progress_with_progress
_advanced.isdn
.send_progress
_as_ alerting
Tones.net.ring _advanced.sip.
progress_if_media
Result
Alert message with
Progess Indicator
0: do not pass media
through
1: pass through
media if in-band
media indicated
2: assume media is
present and pass it
through even if not
indicated in signalling
Progress message with
Progress Indicator
0: do not pass media
through
1: pass through media if in-
band media indicated
2: assume media is present
and pass it through even if
not indicated in signalling
Treat an
incoming ISDN
progress
message as
though it were
an Alerting
message.
Generate ringback tone to
packet network if Alerting or
Progress is received,
provided that no media is
indicated.
0: Force use of180 if
alerting
1: Use 183 rather
than 180 if media
present in original
ISDN alerting or
progress message
2: Use 183 if either
in-band media or
locally generated
media is present
0 180 (no
sdp)
0
1, 2 183 (no
sdp)
0, 1 180
with sdp;
Generated
ringback
0, 1 X X
1
2 183 with
sdp;
Generated
ringback
Alerting
(no media)
2 As Alerting (with media)
0 As Alerting (no media)
0 180
with sdp;
ISDN media
Alerting
(with
media)
1, 2 X X X
1, 2 183
with sdp;
ISDN media
0 X
180
(no sdp)
0,1 180
with sdp;
Generated
ringback
0
1
2 183
with sdp;
Generated
ringback
0, 1
1 As Alerting (no media)
Progress
(no media
indicated)
X
2 As Progress (with media)
0 0 X X 180
(no sdp)
1, 2 0 X X 183
with sdp;
ISDN media
0
X 1 As Alerting (no media)
0 0 X X 180
(no sdp)
1, 2 0 X X 183
with sdp;
ISDN media
Progress
(with
media)
1, 2
X 1 As Alerting (with media)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 261 - 6/2/2009
13.6.3.7 CAS SIP parameters for ringback generation to the VoIP interface
The following table shows the interaction of various parameters with the generation of ringback
tone to the SIP interface.
On setting up a call, after the CAS dialling is complete the Vega CAS code sends a progress
message with no media indication to SIP.
e1t1.port.x.rbs.progress_tones_present tones.net.ring Result
0: Indicate no progress tone
1: Indicate progress tone
Generate ringback tone to packet network when
Alerting or Progress is received, provided that no media
is indicated.
0 180
(no sdp)
0
1 183
with sdp:
Generated ringback
1 x 183
with sdp:
CAS media
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 262 - 6/2/2009
13.7 Symmetric RTP / Dynamic RTP
Symmetric RTP / Dynamic RTP allows the Vega to be configured so that it monitors the incoming
audio RTP stream for a call and makes sure that the RTP it sends out is sent back to that same IP
address as the media is received from. This helps traverse firewalls where the sender does not
properly define the outside IP address of the firewall in its SIP sdp.
Receiving RTP audio data from an IP port and / or IP address that is different from that indicated in
the SDP is not a problem for the Vega receiving the RTP traffic. If however the Vega sends its
RTP traffic back to the originator using the IP address / IP port specified in the SDP it is unlikely to
get through the NAT as the NAT will only route data back to the sender if it is received on the same
IP address / IP port that the RTP traffic is sent from.
In order to handle this, it is necessary for the Vega receiving the RTP to detect the IP port / IP
address that it is receiving the RTP traffic from and return the RTP traffic back to that IP port / IP
address.
[media.control.1.dynamic_update]
enable=1 ; enable
frequency=n ; a value of 0 means that only the first received RTP
packet will be checked. A value of 1 means that every
packet will be checked, a value of 2 means that every
other packet will be checked
ip_follow=1 ; set to 1 to allow IP address and IP port following
private_subnet_list_index=0 ; defines list of allowable IP addresses to follow
If Symmetric RTP is needed, audio cannot be received by the
device whose RTP is being NATed differently from that defined
in the SDP, until the far end has received RTP traffic from that
device (as it is not until the RTP traffic is received that the
returned RTP traffic can be sent to the correct IP port / IP
address). This means that early audio may be lost as initially
it will be sent to the wrong destination IP port / IP address (the
IP port / IP address specified in the SDP).
WARNING!
Checking every packet for a change of IP details is processor
intensi ve benchmark your system if you set
dynamic_update_freq to anything other than zero
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 263 - 6/2/2009
14 FAX, MODEM AND DATA CALLS
14.1 Fax and Modem Operation
In the same way that DTMF tones can be compressed so much that when uncompressed they are
out of specification, so can group 3 fax and modem transmissions. This causes fax / modem tone
recognition problems and therefore failed fax / modem calls.
Vega gateways support both T.38 and G.711 up-speeding to allow fax and modem calls to
succeed:
T.38 is an ITU-T standard defining how to carry group 3 fax transmissions as out of band
packets over an IP network (this only supports fax communications, it does not support
direct modem communications).
Super G3 faxes using modem signalling >33 kbps and non-fax modems require
connection via G.711.
Call flow:
Vega gateways will always connect initially using the preferred voice codec. If fax or modem
detection is enabled (see below for details) then the Vega will monitor for these in-band tones.
When detected, depending on the configuration of the Vega and the tones heard (modem and fax,
or just modem) the Vega will connect using T.38, or up-speed to a data mode G711 codec.
NOTE
1. As per the standards:
H.323 Vega gateways support both TCP and UDP T.38
SIP Vega gateways support UDP T.38 (SIP Annex D T.38)
and also SIP Annex E (voice and fax codec negotiated so
no re-invite needed)
2. Once switched to T.38 mode the Vega will not automatically
revert back to voice mode (it needs a VoIP request to
change back to a voice codec).
3. Vega gateways support connection rates up to 14.4 kbps
when using T.38 (faster connection rates require G.711
data mode)
For further details on the T.38 protocol see Information Note IN_06-T38 protocol interactions.
WARNING!
If you have problems getting fax / modem communications
working look out for the following Gotchas:
1. Delays introduced by the data network can create
problems with the fax handshaking. This is because,
although tones are detected and regenerated at the VoIP
gateways, the handshaking is passed between the end fax
machines.
2. If the clocking of the source and destination VoIP gateways
is not synchronised by say connection of the gateways to
digital trunks on the PSTN, then they will run at
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 264 - 6/2/2009
independent clock speeds. Over time, internal buffers will
overflow or underflow due to the difference in clock (data)
rates. This will cause the fax machines / modems to have
to re-negotiate. If the slip is too great then re-negotiation
will take more time than data transmission time and
connections are likely to fail.
14.1.1 SIP handling of Fax and modem calls
Fax machines and modems only send tones once a call is in progress, so initially a VoIP call will be
set up using a codec specified in medi a. capset . x. caps. If fax and modem detection is
enabled the Vega will then monitor for fax and modem tones. If they are detected, the Vega will do
its best to get the fax / modem call through to the destination, by using either T.38, if enabled, and
if it is supported by the other endpoint device (and the call is a a fax call), otherwise using a G.711
data codec (g.723.1 and G.729 will not pass fax or modem calls).
On detecting the fax tones the Vega first sends a Re-INVITE to the other SIP device with T.38 in
the SDP. If the other end cannot support T.38 then it will reject this Re-INVITE and the Vega will
send another Re-INVITE, this time offering to use G.711U-law and G.711A-law.
If both Re-INVITE's are rejected then the call will be terminated.
If the call is a modem call the INVITE with T.38 will be omitted.
If SIP Annex E is enabled (si p. t 38_annexe_use / si p. t 38_annexe_accept ) and agreed
during sdp negotiation, then the re-invite stage is omitted; when the fax call is detected the media
can be swapped to T.38 immediately.
Some endpoints are sensitive to the SIP header information supplied when making T.38
connections if problems occur, try making the following Vega parameter changes:
[_advanced.sip.sdp]
sess_desc_connect i on=1
t 38_si ngl e_medi a=1
Some fax machines have integrated phone handsets. If a voice call is made between two
such machines (and the call is routed via a VegaStream gateway over SIP), then a FAX is sent
on the same call; if the handsets remain off-hook the two parties can talk to one another again
after the FAX call has been sent.
This will result in the Vega transmitting a further SIP re-INVITE to switch back to a voice codec.
For more details on the operation of the T.38 protocol see IN_06-T38 protocol interactions.
14.1.2 H.323 handling of Fax and modem calls
Fax machines and modems only send tones once a call is in progress, so initially a VoIP call will be
set up using a codec specified in the medi a. capset . x. caps. Typically this capset will be the
faststart capset and will not include any fax or modem codecs. If the Vega detects any fax /
modem tones and the non-faststart capset includes any fax / modem handling codecs, the Vega
will do its best to get the fax / modem call through to the destination, by using either T.38 (tcp or
udp whichever is enabled), if it is supported by the other endpoint device (and the call is a a fax
call), otherwise using a G.711 data codec (g.723.1 and G.729 will not pass fax or modem calls).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 265 - 6/2/2009
On detecting fax or modem tones the Vega closes the voice logical channel and starts media
negotiations to open the relevant T.38 and / or G.711 logical channel (whichever is included in the
non-faststart capset).
If this new media negotiation fails then the call will be terminated.
Some gateways (like Vega gateways) allow T.38 to be included in the original faststart. It is
possible that both a voice and a T.38 channel will be accepted. Under this condition, there is no
need to re-negotiate codecs when fax is detected, fax media will just be sent down the T.38 logical
channel, and voice media will no longer be sent down the voice channel when fax is detected.
NOTE
When using T.38 use of f ast _st ar t is not mandatory, in
fact VegaStreams recommended configuration is to enable
ear l y_h245 and disable f ast _st ar t
For more details on the operation of the T.38 protocol see IN_06-T38 protocol interactions.
14.2 Configuration Parameters for fax / modem handling
[sip]
enabl e_modem=1 ; Al l ow l ow speed modems t o be det ect ed and
; up- speed t o G. 711 i nst ead of usi ng T. 38
f ax_det ect =t er mi nat i ng ; At whi ch end of t he VoI P l i nk shoul d f ax
; t ones be l ooked f or
modem_det ect =t er mi nat i ng ; At whi ch end of t he VoI P l i nk shoul d
; modemt ones be l ooked f or
T38_annexe_accept =0 ; Accept T. 38 Annex E r equest s
T38_annexe_use=0 ; I ni t i at e T. 38 Annex E r equest s
[dsp.t38]
cd_t hr eshol d=- 33 ; Thr eshol d f or Car r i er Det ect si gnal ( db)
FP_FI FO_nom_del ay=300 ; Fax Pl ay- out FI FO nomi nal del ay ( ms)
net wor k_t i meout =150 ; Ti me bef or e cl ear down i f packet s st op
packet _t i me=40 ; Packet si ze i n mi l l i seconds
r at e_max=144 ; Max f ax r at e bps/ 100
r at e_mi n=24 ; Mi n f ax r at e bps/ 100
r at e_st ep=24 ; St ep si ze i n f ax r at es
t i meout =15 ; No Act i vi t y t i meout
t x_l evel =- 8 ; Fax ModemTr ansmi t Level ( 0: - 13dB)
[media.packet.t38tcp.x]
max_r at e=144 ; Pr ef er r ed max f ax r at e bps/ 100
t cf =l ocal ; T. 38 f ax t r ai ni ng mode
[media.packet.t38udp.x]
max_r at e=144 ; Pr ef er r ed max f ax r at e bps/ 100
t cf =t r ansf er r ed ; T. 38 f ax t r ai ni ng mode
[_advanced.dsp]
f ax_di sconnect _del ay ; Del ay af t er r ecei vi ng di sconnect bef or e
; cl ear i ng cal l
t 38_di ags=0 ; For engi neer i ng use onl y
[_advanced.dsp.buffering.fax]
dept h=100 ; Buf f er si ze
enabl e=0 ; Enabl e T. 38 packet r e- synch i n buf f er
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 266 - 6/2/2009
[_advanced.media]
cont r ol _v25 = f ax ; For ce t o f ax mode i f V25 t one i s hear d
[_advanced.t38]
al l ow_MR_page_compr ess=1 ; Do not suppr ess use of MR page
; compr essi on
al l ow_ecm=1 ; Do not suppr ess Er r or Cor r ect i on Mode
enabl e_Ef l ags_i n_f i r st _DI S=1 ; For Engi neer i ng use onl y
enabl e_TFoP=1 ; Do not di sabl e r epet i t i on of
; Fr ameCompl et e packet
enabl e_scan_l i ne_f i x_up=1 ; Do not di sabl e scan l i ne f i x- up
[_advanced.t38.tcp] (H323 Only)
col l ect _hdl c ; Col l ect V. 21 hdl c i nt o packet s
connect _on_demand=1 ; Connect T. 38 when f ax t ones ar e det ect ed
; ( r at her t han on ever y cal l )
por t _r ange_l i st =2 ; _advanced. l an. por t _r ange_l i st t hat
; speci f i es t 38 t cp por t s
suppr ess_t 30=0 ; Suppr ess t r ansmi ssi on of some T. 30
; i ndi cat i ons
[_advanced.t38.udp]
check_st ar t _packet =1 ; Onl y swi t ch t o f ax mode when f i r st f ax
; packet has been r ecei ved ( al l owi ng voi ce
; pat h t o r emai n connect ed t o t hat poi nt )
por t _r ange_l i st =3 ; _advanced. l an. por t _r ange_l i st t hat
; speci f i es t 38 udp por t s
H.323 Vega gateways treat TCP T.38 and UDP T.38 as codec types. Enabling T.38 is carried out
in the same manner as enabling audio codecs; see section 13.1.4 Defining Fax capabilities.
SIP gateways treat UDP T.38 as a codec type. Enabling T.38 is carried out in the same manner as
enabling audio codecs; see section 13.1.4 Defining Fax capabilities.
More details on some of the key parameters:
[media.packet.t38tcp.x] (H323 only)
t cf
The t cf parameter defines whether fax modem training is carried out at the local ends of the VoIP
link, or whether the training tones should be transferred across the VoIP link for t38 tcp
recommendations say keep training local
It is important that this value is configured the same at both ends of the VoIP call.
[media.packet.t38udp.x]
t cf
The t cf parameter defines whether fax modem training is carried out at the local ends of the VoIP
link, or whether the training tones should be transferred across the VoIP link for t38 udp
recommendations say transfer the training information across the VoIP link
It is important that this value is configured the same at both ends of the VoIP call.
[sip]
enabl e_modem
If enabl e_modemis set to 0, then the Vega will not support low speed modems; it will treat any
call which has low speed modem tones as a fax call. This setting can be used if it is known that all
calls will be voice or fax calls and not modem calls.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 267 - 6/2/2009
If enabl e_modemis set to 1, then, on hearing low speed modem tones, the Vega will assume that
the call is a low speed modem call (and use G.711 rather than T.38) unless it detects the V.21 tone
which confirms that the call is a fax call.
If enabl e_modemis set to 1, then even if G711 data codecs are not enabled in the active
medi a. capset . n. caps they may still be used.
[sip]
f ax_det ect
modem_det ect
The f ax_det ect and modem_det ect parameters defines whether the Vega looks for fax and /
or modem tones: only from its telephony interface, from telephony and VoIP interfaces, or never.
It is generally better (and adheres to the standards) to only detect tones on one end of the call
the end terminating the VoIP call (initiating the call to the answering fax machine / modem); this is
the end that hears the tones directly from the line (rather than having to detect tones that have
passed through both the telephone line and through VoIP). If the far end 3
rd
party gateway does
not detect the tones properly the Vega can be configured always to detect fax / modem tones,
whether the call arrives on the Vega on its telephony interface or its VoIP interface.
[sip]
T38_annexe_accept
T38_annexe_use
T.38 Annex E allows media to change from Voice to T.38 without need for a re-invite. This speeds
up the transition from voice mode to fax mode, and reduces the number of signalling messages
required.
[_advanced.media]
cont r ol _v25
Setting v25_cont r ol to data causes the Vega to use G711 data codecs rather than T.38 for
transmission of modem and fax calls.
[_advanced.dsp.buffering.fax]
dept h ; Buf f er si ze
enabl e ; Enabl e T. 38 packet r e- synch i n buf f er
By default Vega gateways expect to see T.30 / T.38 messages arriving in sequence. With certain
gateways (e.g. Cisco) the messages are not always sent out sequentially. By enabling
_advanced. dsp. buf f er i ng. f ax the Vega can handle this. It re-orders the T.30 / T.38
messages into sequential order as it puts them in the buffer.
For details about other parameters, see the information in 6.7 Configuration Entries, and
6.8 Advanced configuration entries.
14.2.1 Recommended Values For SIP FAX / Modem Connecti vity
For normal use with FAX and modems:
1. Enable the required audio codecs in the capset. Add T38udp, followed by one or both
G711Alaw 64k profile 2 and / or G711ulaw64k profile 2.
2. set si p. enabl e_modem= 1
3. set _advanced. medi a. cont r ol _v25 = i gnor e
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 268 - 6/2/2009
For use with G.711 Up-Speeding only, and no T.38:
1. Enable the required audio codecs in the capset. Add one or both G711Alaw 64k profile
2 and / or G711ulaw64k profile 2.
2. set si p. enabl e_modem= 1
3. set _advanced. medi a. cont r ol _v25 = dat a
For use with T.38 only, and no G711 Up-Speeding:
1. Enable the required audio codecs in the capset, add T38udp as the last entry.
2. set si p. enabl e_modem= 0
3. set _advanced. medi a. cont r ol _v25 = f ax
14.3 ISDN Unrestricted Digital Information Bearer Capability And Clear Mode
ISDN calls calls are tagged with a bearer capability identifying the type of media being carried. For
standard Voice and fax calls, bearer capabilities of voice and 3.1KHz audio are usually used.
One of the other bearer capabilities is Unrestricted Digital Information. In order to carry this type of
media, standard voice compression / gain must not be applied. SIP variants of Vega code
automatically force the codec type to use to clear mode when Unrestrictd Digital Information calls
are received.
Clear mode (also known as octet codec) can also be specified in the capset to be used in cases
where a bearer capability is not available or the one received does not specify Unrestricted Digital
Information.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 269 - 6/2/2009
15 SIP GATEWAYS
This section describes the configuration and behaviour of SIP variants of the Vega gateway.
15.1 Introduction
The SIP firmware acts as a set of SIP User Agents within the Vega. Communication, by default, is
via UDP unicast, usually to and from UDP port 5060. TCP connection for SIP signalling messages
may also be configured. (Note audio RTP traffic is always UDP).
All Request URI usernames are of the form si p: t el ephone_number and all hosts are
expressed as numerical IP addresses, or domain names if DNS is configured, in which case
l an. name must be set to the Vegas DNS hostname.
The SIP module supports remote commands for re-INVITE, hold and retrieve, transfer via the BYE-
Also mechanism and also the REFER method.
Calls are accepted either solely from a designated default proxy (or from its backups), or from any
source, depending on a configuration option.
Calls are routed between the telephony interfaces and the SIP module by means of dial plans. The
SIP module being represented by the default interface ID of 99.
The module may be configured to provide reliable provisional responses (PRACK) when receiving
the Require: or Supported: headers. The module may also be configured to request reliable
provisional responses using the Require:100rel or Supported:100rel.
For FXS units, the SIP module also includes mechanisms for handling Flash-hook, DTMF, call
waiting, message waiting and distinctive ringing.
Vegas also feature the ability to generate tones toward the network and an off-hook warning tone
towards a phone.
All Vega gateways may be configured to register with a registration server (typically part of the
proxy).
All Vega gateways also support Authentication on Registration, INVITE, ACK and BYE messages.
15.2 Monitor Commands
SI P MONI TOR ON
SI P MONI TOR OFF
Control the display of the SIP signalling monitor. The monitor is useful for checking the operation of
the SIP module. The monitor displays each SIP message sent or received, headed by an output
line in the following form:
SI P m: Syst em_el apsed_t i me( ms) del t a_t i me( ms) message_number <- - RX/ TX - - - Fr om/ To
I P_addr ess: Por t
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 270 - 6/2/2009
15.3 Registration Status Commands
Registration is supported on all Vega gateways.
Please Refer to sections 15.4.4 Registration, and 15.4.5 Authentication for setup details.
By default Vega gateways are configured not to register by default, but FXS ports and FXO ports
have registration entries configured and disabled so that they are easy to enable.
The console registration status and registration commands are:
* SI P SHOWREG
* SI P SHOWREG [ user ]
* SI P REG user
* SI P REG ALL
* SI P CANCEL REG user
* SI P CANCEL REG ALL
* SI P RESET REG
15.3.1 SIP SHOW REG
Use this command to display the current registration state of all SIP registration users.
Syntax SIP SHOW REG
Behaviour: Displays the current registration state of ALL records as in the following example:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
domai n = abcdef ghi j what ever . com
expi r y = 600
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
SI P REG USER 1
- - - addr ess - si p: 01@abcdef ghi j what ever . com
- - - aut h user aut h user di sabl ed
- - - cont act - si p: 01@172. 16. 30. 31
- - - st at e - r egi st er ed
- - - TTL - 500 seconds
SI P REG USER 2
- - - addr ess - si p: 02@abcdef ghi j what ever . com
- - - aut h user aut h user di sabl ed
- - - cont act - si p: 02@172. 16. 30. 31
- - - st at e - r egi st er ed
- - - TTL - 480 seconds
. . .
15.3.2 SIP SHOW REG [user]
Syntax SIP SHOW REG [user]
user optional parameter to specify which user's details you wish to see.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 271 - 6/2/2009
Example SIP SHOW REG 1
Behaviour Vega displays the registration status of the users / all users
15.3.3 SIP REG user
Syntax SI P REG user
Example SI P REG 1
Behaviour Vega sends a "register user" message to the registration server for the specified user.
15.3.4 SIP REG ALL
Syntax SI P REG ALL
Behaviour Vega sends "register user" messages to the registration server for ALL users.
15.3.5 SIP CANCEL REG user
Syntax SI P CANCEL REG user
Example SI P CANCEL REG 1
Behaviour Vega sends a "cancel registration" message to the registration server for the specified user.
15.3.6 SIP CANCEL REG ALL
Syntax SIP CANCEL REG ALL
Behaviour Vega sends "cancel registration" messages to the registration server for ALL users.
15.3.7 SIP RESET REG
Syntax SIP RESET REG
Behaviour Vega cancels all current registrations and re-registers the updated user details with the
registration server (used on re-configuration of registration details).
15.4 SIP Configuration
SIP configuration is performed in the SIP subsection of the configuration database. This can be
accessed via a web browser or via the command line interface. The following notes refer to the
command line interface procedures.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 272 - 6/2/2009
15.4.1 TCP / UDP SIP
The Vega can be configured to send SIP signalling messages using UDP, TCP or TLS. This is
configurable on a per SIP profile basis:
[sip.profile.1]
sig_transport=udp ; udp, tcp or tls
UDP has been part of the SIP standards for longer, and so if the Vega is configured for TCP
operation and it cannot get a TCP connection it will revert back to UDP for that call.
SIP over TLS is an optional addition to the standard featureset and requires a special license to
enable. TLS (Transport Layer Security) secures the TCP/IP connection and hence secures the
SIP messaging.
15.4.2 Proxy
Vega gateways can be configured to operate with SIP Proxy servers. This is a common
configuration, especially where advanced features, like follow me, conferencing or voice mail are
required. Also where centralisation of the configuration of routing data is required, or connection to
an ITSP (Internet Telephony Service Provider) is required.
A proxy sever is a device to which the Vega can send SIP call traffic.
The parameter si p. pr of i l e. x. pr oxy. y. i pname is used to define the IP address of the proxy
server that you wish the Vega to communicate with (i.e. where to send the INVITE messages to).
The proxy IP address may be defined either as a dotted decimal value, e.g. aaa. bbb. ccc. ddd or:
as a DNS name, e.g. sip.vegastream.com
NOTE
If SIP calls are to be sent to destinations other than the Proxy,
the TA: token in the dial planner can be used to override the
destination IP address.
15.4.2.1 Multiple SIP Proxy Support
Vega gateways support the ability to use more than 1 proxy for redundancy and for load balancing
purposes. Either multiple alternative SIP proxies can be defined through use of a list of proxies, or
multiple alternative SIP proxies can be defined through use of DNS SRV records on a single DNS
SRV name.
15.4.2.1.1 Multiple SIP Proxy Configuration
The configuration parameters used in "multiple proxy support" are:
[sip.profile.x.proxy]
mi n_val i d_r esponse=180 ; Once the Vega receives a SIP message
response whose ID >=value specified by this
parameter, it knows that the proxy is "up" and the
Vega will not try other proxies in the list (i.e. any
SIP responses with a value less than
"min_valid_response" will be ignored by the
"multiple proxy support" module). The exception
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 273 - 6/2/2009
to this rule is when 5xx responses (e.g. "500
internal server error") are received. In such a
case, the INVITE will be sent to the next backup
proxy immediately.
mode=nor mal ; the mode parameter defines how the Vega
should handle the alternate proxies:
normal use the first proxy in list that responds
with a valid response
cyclic for each call try the next proxy in the list
dnssrv abide by the dnssrv proxy list and
weighting (Vega only uses first proxy entry)
t i meout _ms=5000 ; if the Vega does not receive a "minimum valid
response" to an INVITE within the time specified
by this parameter, then the Vega will try the next
proxy in the list.
Determining whether the proxy is available to use
accessi bi l i t y_check=of f ; off: Only treat proxy as failed if SIP timeouts fail
the call then use alternate proxy for that call
options: Treat proxy as failed if SIP OPTIONS
messages are not responded to by the proxy
(use alternate proxy for all calls until OPTIONS
messages are responded to again)
BYE: Same behaviour as options but uses SIP
BYE messages to poll proxy availability.
accessi bi l i t y_check_t r anspor t =udp ; Specify whether to use udp or tcp to send
OPTIONS messages to the proxy (to see if it is
alive)
r et r y_del ay=0 ; When a proxy is deemed to have failed and the
Vega switches to using an alternate proxy, this
timer specifies how long to wait before trying that
failed proxy again (allowing the proxy time to
recover and minimising the delay on future
phone calls they do not have to time out
before being routed using the alternative proxy)
[sip.profile.x.proxy.1] ; first proxy / DNS SRV name
enabl e=1 ; 1 =enable requests to this proxy
i pname=136. 170. 208. 134 ; the IP address or resolvable DNS name of the
backup proxy
por t =5060 ; the port to use for this proxy (not used when
mode =dnssrv as dnssrv supplies IP port)
[sip.profile.x.proxy.2] ; second proxy
enabl e=1
i pname=hel l o. com
por t =5060
[sip.profile.x.proxy.3] ; third proxy
et c
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 274 - 6/2/2009
NOTE
The default value chosen for mi n_val i d_r esponse is 180
(ringing) because it means that the call is REALLY
progressing.
A value of say 100 (trying) could be used this would
indicate that the proxy is alive, but it only indicates that the
proxy received the message - it doesn't necessarily mean
that the proxy has done anything useful with it.
NOTE
Configuring a registrar and alternatives follows the same
methodology as configuring the proxy and alternatives
15.4.2.1.2 Commands associated with Multiple SIP proxies
new si p. pr of i l e. x. pr oxy
Adds a new entry
del et e si p. pr of i l e. x. pr oxy. n
Deletes an entry
NOTE
You can only delete the last backup_proxy.n in the
backup_proxy list.
15.4.2.1.3 Examples of Multiple Proxy Support Operation Normal mode
1. Single proxy operation
Vega simply sends INVITE to the default proxy e.g.:
Vega- - - - I NVI TE- - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
2. Operation with two proxies
Vega starts by sending the INVITE to the default proxy e.g.:
Vega- - - - I NVI TE- - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
If the default proxy does not respond with at least a mi n_val i d_r esponse (typically=180)
message within backup_pr oxy. t i meout _ms (e.g. 5000ms) then the Vega will send out a new
INVITE to the second proxy.
Vega- - - - I NVI TE- - - - >136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
If the second proxy responds with at least a mi n_val i d_r esponse message within
backup_pr oxy. t i meout _ms then the Vega will try to cancel the INVITE to the default proxy.
Vega<- - - - 100 Tr yi ng- - - - 136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
Vega<- - - - 180 Ri ngi ng- - - - 136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 275 - 6/2/2009
Vega- - - - - CANCEL- - - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
3. Operation with three proxies
Vega starts by sending the INVITE to the default proxy e.g.:
Vega- - - - I NVI TE- - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
If the default proxy does not respond with at least a mi n_val i d_r esponse (typically=180)
message within backup_pr oxy. t i meout _ms (e.g. 5000ms) then the Vega will send out a new
INVITE to the second proxy.
Vega- - - - I NVI TE- - - - >136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
If the second proxy also does not respond within backup_pr oxy. t i meout _ms, then the Vega
will send out a new INVITE to the third proxy.
Vega- - - - - - - I NVI TE- - - - - - >136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
If the third proxy responds with at least a mi n_val i d_r esponse message within
backup_pr oxy. t i meout _ms then the Vega will try to cancel the INVITE to the default proxy and
second proxies.
Vega<- - - - 100 Tr yi ng- - - - 136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
Vega<- - - - 180 Ri ngi ng- - - - 136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
Vega- - - - - CANCEL- - - - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
Vega- - - - - CANCEL- - - - - - >136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
4. Operation with three proxies (2
nd
proxy returns with a server error)
Vega starts by sending the INVITE to the default proxy e.g.:
Vega- - - - I NVI TE- - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
If the default proxy does not respond with at least a mi n_val i d_r esponse (typically=180)
message within backup_pr oxy. t i meout _ms (e.g. 5000ms) then the Vega will send out a new
INVITE to the second proxy.
Vega- - - - I NVI TE- - - - >136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
If the second proxy responds with a server error, then the Vega sends a new INVITE to the third
proxy (immediately not waiting the backup_pr oxy. t i meout _ms delay).
Vega<- - 501 Ser ver Er r or - - 136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
Vega- - - - - ACK- - - - - - - - >136. 170. 208. 134 (sip.profile.x.proxy.2.ipname)
Vega- - - - I NVI TE- - - - >136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 276 - 6/2/2009
Once the proxy responds with a 180 message the Vega will tries to cancel any other outstanding
INVITE.
Vega<- - - - 100 Tr yi ng- - - - 136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
Vega<- - - - 180 Ri ngi ng- - - - 136. 170. 208. 200 (sip.profile.x.proxy.3.ipname)
Vega- - - - - CANCEL- - - - - - >136. 170. 208. 133 (sip.profile.x.proxy.1.ipname)
The Vega does not need to CANCEL the INVITE to the second proxy because the transaction has
already been completed with the "501 Server Error" and ACK response
15.4.2.1.4 Examples of Multiple Proxy Support Operation Cyclic mode
If
[sip.profile.x.proxy.1]
def aul t _pr oxy=200. 100. 50. 1
[sip.profile.x.proxy.2]
enabl e=1
i pname=200. 100. 50. 2
[sip.profile.x.proxy.3]
enabl e=1
i pname=200. 100. 50. 3
on the first call after power-up, the Vega would try the SIP proxy at 200.100.50.1 and then, if
there was no response, 200.100.50.2, and then 200.100.50.3.
On the second call, the Vega would first try the SIP proxy at 200.100.50.2 (the 2
nd
proxy) and
then, if there was no response, 200.100.50.3, and then 200.100.50.1.
Then, on the third call, the Vega would first try the SIP proxy at 200.100.50.3 (the 3
rd
proxy) and,
if there was no response, 200.100.50.1, and then 200.100.50.2.
And on the fourth call 4, the Vega would start again with the default proxy (as per the first call).
This "cyclic" mode provides a primitive form of load-balancing of calls over the listed proxies.
15.4.3 SIP SDP a= ptime and direction attributes
15.4.3.1 Ptime attribute in SDP
In SIP SDPs a codec Packet Time (ptime) may be requested / specified. Control over whether the
Vega will ignore and not generate ptime requests, or whether it will act upon and generate ptime
parameters is controlled by the parameter:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 277 - 6/2/2009
[_advanced.sip.sdp]
pt i me_mode ; 0=i gnor e / do not gener at e pt i me,
; 1=act upon and gener at e pt i me
; mpt i me
; x_mpt i me
; pt i me 30
; pt i me60
If pt i me_mode=0 then the Vega will neither create, nor respond to ptime requests.
If pt i me_mode=1 then the Vega will create and respond to ptime requests based on its codec
capabilities.
Vegas support the following codecs and packet times:
G.729 - 10, 20, 30, 40, 50, 60, 70 or 80ms
G.711a - 10, 20 or 30ms
G.711u - 10, 20 or 30ms
G.723.1 - 30 or 60ms
1) If the Vega receives an INVITE including a codec and ptime that it supports, it will honour the
ptime and respond with that codec and the ptime in its returning the SDP
For example:
<- - I nvi t e:
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
a=pt i me: 20
- - >Ri ngi ng/ OK
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
a=pt i me: 20
2) If the incoming INVITE does not specify the ptime, the Vega will inform the originator of its
choice by supplying the ptime in its SDP.
For example:
<- - I nvi t e:
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
- - >Ri ngi ng/ OK
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
a=pt i me: 30
3) If the Vega cannot honour the requested ptime, it responds with a 488 error (Not Acceptable
Here) and specifies the unsupported ptime.
For example:
<- - I nvi t e:
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
a=pt i me: 950
- - >488 audi o pt i me 950ms unsuppor t ed or unobt ai nabl e
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 278 - 6/2/2009
There will also be a log message:
LOG: 14/ 03/ 2003 09: 56: 43. 660 SI P ( I ) Rd3C00 unsuppor t ed/ unobt ai nabl e
packet t i me ( 950 ms)
cal l r ef =[ f 100001f ]
4) If G723 is requested, the Vega forces a ptime based on the value configured in
medi a. packet . g7231. y. packet _t i me, regardless of the original request.
For example if packet _t i me=30:
<- - I nvi t e:
m=audi o 10000 RTP/ AVP 4 - - - g723
a=pt i me: 20
- - >Ri ngi ng/ OK
m=audi o 10000 RTP/ AVP 4 - - - g723
a=pt i me: 30
5) INVITEs sent by the Vega will specify the ptime as that configured in the
medi a. packet . xxxx. y. packet _t i me configuration parameter. In case where there are
multiple codecs with different packet times being specified, the packet time of the first codec will be
used.
For example, assuming
g723 configured to use 30ms packet time
G.711 u-law configured to use 20ms packet time
- - >I nvi t e:
m=audi o 10000 RTP/ AVP 0 4 - - - G. 711 u- l aw or g723
a=pt i me: 20
<- - Ri ngi ng
m=audi o 10000 RTP/ AVP 0 - - - G. 711 u- l aw
a=pt i me: 20
Or:
- - >I nvi t e:
m=audi o 10000 RTP/ AVP 4 0 - - - g723 or G. 711 u- l aw
a=pt i me: 30
<- - Ri ngi ng
m=audi o 10000 RTP/ AVP 4 - - - g723
a=pt i me: 30
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 279 - 6/2/2009
6) If a Vega gets a ptime in the "SDP answer", the Vega will try to use it if it can. If it cannot, it will
try to hangup the call and then add a message to the log:
For example:
- - >I nvi t e:
m=audi o 10000 RTP/ AVP 4 - - - g723
a=pt i me: 20
<- - Ri ngi ng
m=audi o 10000 RTP/ AVP 4 - - - g723
a=pt i me: 300
- - >Cancel
There will also be a log message:
LOG: 14/ 03/ 2003 09: 56: 43. 660 SI P ( I ) Rd3C00 unsuppor t ed/ unobt ai nabl e
packet t i me ( 300 ms)
cal l r ef =[ f 100001f ]
If pt i me_mode=mpt i me then the Vega will offer a list of ptimes, one for each codec, e.g. the sdp
will look like:
m=audi o 10002 RTP/ AVP 0 8 4 18 96
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=r t pmap: 4 G723/ 8000
a=r t pmap: 18 G729/ 8000
a=r t pmap: 96 t el ephone- event / 8000
a=f mt p: 96 0- 15, 16
a=mpt i me: 30 30 30 20 -
a=sendr ecv
In the above example, the packet time is 30ms G.711u-law, for 30ms for G.711a-law, 30ms for
g723.1 and 20ms for 729. The packet times used correspond to the
medi a. packet . xxx. y. packet _t i me configuration parameters where xxx is the codec and y is
the codec profile; NOTE: a dash is used for the telephone event packet time because the packet
time used for telephone events corresponds to the packet time of the selected codec.
If pt i me_mode=x_mpt i me then the Vega will offer a list of ptimes, one for each codec, just as for
pt i me_mode=mpt i me; in this mode however, the key word is X-mptime: i.e.:
a=X- mpt i me: 30 30 30 20 -
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 280 - 6/2/2009
If pt i me_mode=pt i me30 then the Vega will offer a 30ms value, unless all codecs are G.711,
when it will use a 20ms, e.g. for G.711 codecs:
m=audi o 10002 RTP/ AVP 0 8
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=pt i me: 20
a=sendr ecv
e.g. for codecs which include non G.711 codecs:
m=audi o 10002 RTP/ AVP 0 8 4 18
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=r t pmap: 4 G723/ 8000
a=r t pmap: 18 G729/ 8000
a=pt i me: 30
a=sendr ecv
If pt i me_mode=pt i me60 then the Vega will offer a 60ms value if all offered codecs are capable
of supporting 60ms. If all codecs are G.711, then a value of 20ms will be used, and if not all
codecs are G.711, but 60ms is not supported by all codecs then 30ms will be used.
e.g. for G.711 codecs only:
m=audi o 10002 RTP/ AVP 0 8
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=pt i me: 20
a=sendr ecv
e.g. for all codecs (G.711 does not support 60ms):
m=audi o 10002 RTP/ AVP 0 8 4 18
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=r t pmap: 4 G723/ 8000
a=r t pmap: 18 G729/ 8000
a=pt i me: 30
a=sendr ecv
e.g. for G.723.1 and G.729 codecs (both which support 60ms packets):
m=audi o 10002 RTP/ AVP 4 18
c=I N I P4 136. 170. 209. 134
a=r t pmap: 4 G723/ 8000
a=r t pmap: 18 G729/ 8000
a=pt i me: 60
a=sendr ecv
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 281 - 6/2/2009
15.4.3.2 Maxptime attribute in SDP
In SIP SDPs a codec Maximum Packet Time (maxptime) may be specified. Control over whether
or not the Vega will try to include a maxptime request in sdps depends on the setting of:
[_advanced.sip.sdp]
maxpt i me_enabl e ; 0=do not i ncl ude maxpt i me,
; 1=t r y t o i ncl ude a maxpt i me
For example, if G.711 A law and u law are offered, with a prefered time of 20ms and each has a
max time (dsp.xxx.packet_time_max) of 30, then the sdp will be as follows:
m=audi o 10002 RTP/ AVP 0 8
c=I N I P4 136. 170. 209. 134
a=r t pmap: 0 PCMU/ 8000
a=r t pmap: 8 PCMA/ 8000
a=pt i me: 20
a=maxpt i me: 30
An a=maxpt i me attribute will only be included in an sdp if it does not contradict other attributes,
and if the a=maxpt i me is valid for all offered codecs.
So, for example if the codecs offered are G.711Alaw and G.729, the maxptime value will be the
smaller of dsp. g711Al aw64k. packet _t i me_max and dsp. g729. packet _t i me_max.
However, a=maxpt i me will only be put in to the sdp if it is consistent with a=mpt i me,
a=X- mpt i me or a=pt i me, i.e it does not specify a time smaller then these preferred times.
If the codecs offered are G.711Alaw and G.723.1, and
dsp. g711Al aw64k. packet _t i me_max=20 then an a=maxpt i me will not be included in the sdp
as a maxptime of 20ms is not valid for G.723.1 (the minimum packet size for G.723.1 is 30ms).
15.4.3.3 Direction attribute in SDP
In SIP SDPs a media direction attriburte may be sent / received. The direction attribute takes one
of the following 4 forms:
a=sendr ecv
a=sendonl y
a=r ecvonl y
a=i nact i ve
The way the Vega handles the sending / receiving of this attribute is controlled by:
[_advanced.sip.sdp]
di r ect i on_at t r i but e ; 0=do not i ncl ude/ handl e di r ect i on at r r i but e
; 1=i ncl ude and handl e di r ect i on at t r i but e
If disabled, the Vega will not include the direction attribute in sdps that it generates; it will also
ignore directon attribute requests that it receives.
If enabled, for calls where the Vega is going to send the first sdp (this Vega is going to make the
offer, the other device is going to answer) the Vega will always include a=sendrecv.
For calls where the Vega is going to respond to an incoming sdp (the other device is going to make
the offer, and this Vega is going to answer) the response the Vega will make is as per the following
table:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 282 - 6/2/2009
Recei ved sdp Vegas sdp response Notes
A=sendr ecv a=sendrecv
A=sendonl y a=recvonly Vega mutes media transmission
A=r ecvonl y a=sendonly Vega mutes media reception
A=i nact i ve a=inactive Vega mutes media Tx and Rx
No direction attribute a=sendrecv
15.4.4 Registration Vega 400, Vega BRI, Vega FXS, Vega FXO
Whether the Vega registers or not is controlled on a per unit basis by:
[ si p]
r eg_enabl e=1 ; 0=do not r egi st er , 1 = r egi st er
The domain, hostname or IP address of the registrar is set using:
[ si p. pr of i l e. x]
r eg_domai n=<domai n, host name or I P addr ess>
The lifetime, s seconds, of all registrations for the unit is configured using:
[ si p. pr of i l e. x]
r eg_expi r y=s
Registration requests are sent to the IP address and port number specified in the following
parameters:
[sip.profile.x.registrar.y]
r eg_pr oxy
r eg_r emot e_r x_por t
If si p. r eg_enabl e=1, then:
[sip]
r eg_on_st ar t up=0 or 1
controls whether the Vega will automatically register on start-up. If si p. r eg_on_st ar t up=0
then registrations will only occur when the first call is made from that port. If
si p. r eg_on_st ar t up=1 then registrations will occur for all enabled registration users on
system power-up or re-boot.
A number of SIP Registration Users may be set up. The parameters to do this are:
[sip.reg.user.1]
aut h_user _i ndex=1
dn=100
enabl e=1
user name=RegUser 1
[sip.reg.user.2]
et c
etc
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 283 - 6/2/2009
The Vega will register with each sip.reg.user.x that is enabled. username forms the central
part of the username used for registration. Dn provides the telephone number part of the contact
information, i.e. dn@ip_address_of_vega.
If the registration server is going to request authentication, then configure auth_user_index to
point to the sip.auth.user.n info that should be used to respond to the authentication
challenge.
NOTE
1. Vega gateways support the ability to use more than 1
registrar for redundancy and for load balancing purposes.
Either multiple alternative Registrars can be defined through
use of a list of Registrars, or multiple alternative Registrars
can be defined through use of DNS SRV records on a single
DNS SRV name.
This operates exactly the same way that Multiple SIP proxies
do see section 15.4.2.1 Multiple SIP Proxy Support for
details.
2. Vega gateways can register with multiple proxies
simultaneously (one per sip profile). For more details see
Using_multiple registrations_on_R8_x_01 on the technical
documents page of www.VegaAssist.com
For more details on the structure of registration and other SIP messages, see IN_10- Introduction
to Vega SIP mesaging.
Also see the SIP REGISTRATON and SIP INVITE configuration utility on the
www.VegaAssist.com (Documentation > Step by step configuration).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 284 - 6/2/2009
15.4.5 Authentication Vega 400, Vega BRI, Vega FXS, Vega FXO
Vega gateways may be configured to respond appropriately to authentication challenges (e.g. to
REGISTRATION, INVITE, ACK and BYE messages).
Vega gateways support the ability to define one or more authentication username and password
combinations to respond to the authentication challenges. The parameters used are:
[sip.auth.user.1]
enabl e=1
user name=aut huser 1
passwor d=pass1
sr ce=I F: 01
[sip.auth.user.2]
enabl e=1
user name=aut huser 2
passwor d=pass2
sr ce=I F: 02, TEL: 0123. *
The username used in the response to the authentication challenge is sip.auth.user.n.username
The username / password combination defined for a user is valid for calls whose telephony details
match the sr ce specification. sr ce can contain the IF: and TEL: tokens to match against the call
details. For telephony to LAN calls, sr ce is matched against the incoming call details, for LAN to
telephony, sr ce is matched against the call details used for making the telephony call (i.e. the
destination call details).
NOTE
1. sr ce may only use Dial Plan srce wildcards, e.g. . * ? [xyz]
it may not use destination wildcards like <1>as this will
not be defined.
2. If the case where different users sr ce expressions
overlap, the Vega will just use the username / password in
the first found user that matches.
15.4.6 Incoming INVITEs
[sip]
accept _non_pr oxy_i nvi t es=0 or 1
controls whether the Vega will accept INVITES from sources other than the configured
default_proxy (and backup proxies).
15.4.7 Local and Remote Rx Ports
The default UDP port number used for SIP signalling is 5060. Sometimes, however, use of a
different port number may be desired.
[sip]
l ocal _r x_por t =1 t o 65535 ; def aul t =5060
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 285 - 6/2/2009
sets the UDP port on which the gateway expects to receive SIP messages. If the value is
non-5060 then the gateway will listen on both ports 5060 and the one specified by
si p. l ocal _r x_por t .
[sip]
r emot e_r x_por t =1 t o 65535 ; def aul t =5060
sets the UDP port to which the gateway should send SIP messages.
15.4.8 PRACK Support
Allows configuration of the gateway to send PRACKs (Provisional ACKnowledgements). By
default this is off but you can set it to supported or required:
[sip]
pr ack=suppor t ed
Permitted values:
of f PRACK not supported at all
suppor t ed the gateway will use PRACK if the remote proxy or gateway requires it
r equi r ed the gateway will insist that the remote proxy or gateway uses PRACK
otherwise the connection will not proceed
15.4.9 REFER/REPLACES
All Vega gateways will respond to the REFER / REPLACES method for transferring calls, but only
FXS gateways can initiate call transfers (initiated using hookflash if supplementary services is
enabled)
9
.
On receiving a REFER, the Vega will send an INVITE (with the replaces header) to the destination
specified in the REFER. If the INVITE resulting from the REFER should be sent via the SIP proxy,
set:
[_advanced.sip]
r ef er _i nvi t e_t o_pr oxy=1
9
See the FXS Call Transfer documnt for more details on configuring FXS ports to initiate call
transfers.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 286 - 6/2/2009
15.4.10 RPID Remote Party ID header
SIP Vegas support the generation and reception of the SIP RPID (Remote Party ID) header in
INVITE messages.
RPID headers provide the SIP recipient with details of the calling party and the original called
number or the (last) redirecting number.
To enable the generation and reception of RPID headers, set:
[_advanced.sip.privacy]
st andar d=r pi d ; def aul t =r f c3323
15.4.10.1 Mapping ISDN SETUP Information Elements to SIP RPID header parameters
Four cases are illustrated to demonstrate the methodology used in translating the paramterters
Case 1 Calling number presentation allowed
ISDN SETUP
10
SIP INVITE
Called party number IE>number digits Request-URI & user part of To:
Calling party number IE>number digits User part of From:
Calling party number IE>presentation (allowed) not explicitly forwarded
Display IE Name part of From:
Case 2 Calling number presentation allowed with original called number or redirection IE
ISDN SETUP SIP INVITE
Called party number IE>number digits Request-URI & user part of To:
Calling party number IE>number digits User part of From:
Calling party number IE>presentation (allowed) not explicitly forwarded
Display IE Name part of From:
Original called number / redirection IE RPID>party=redirect
Original called number / redirection IE>number digits RPID>user=
Original called number / redirection IE>screening
indicator
RPID>screen=
Original called number / redirection IE>Presentation RPID>privacy=
RPID header format:
Remot e- Par t y- I D: r pi d_di sp_name <si p: r pi d_CgPN@domai n; user =phone>; r pi d_opt i ons
e.g.:
Remot e- Par t y- I D: J ohn Smi t h <si p: 01344123456@vegast r eam. com; user =phone>; scr een=yes; par t y=cal l i ng
10
IE stands for Information Element; a message element in ISDN signalling
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 287 - 6/2/2009
Case 3 Calling number presentation restricted
ISDN SETUP SIP INVITE
Called party number IE>number digits Request-URI & user part of To:
User part of From: =restricted user
Name part of From: =restricted
name
Calling party number IE RPID>party=calling
Calling party number IE>number digits RPID>user=
Calling party number IE>Screening indicator RPID>screen=
Calling party number IE>presentation (restricted) RPID>privacy=full
Display IE RPID>display-name
Case 4 Calling number presentation restricted with original called number or redirection
IE
ISDN SETUP SIP INVITE
Called party number IE>number digits Request-URI & user part of To:
User part of From: =restricted user
Name part of From: =restricted
name
Calling party number IE RPID>party=calling
Calling party number IE>number digits RPID>user=
Calling party number IE>screening indicator RPID>screen=
Calling party number IE>presentation (restricted) RPID>privacy=full
Display IE RPID>display-name
Original called number / redirection IE RPID>party=redirect
Original called number / redirection IE>number digits RPID>user=
Original called number / redirection IE>screening
indicator
RPID>screen=
Original called number / redirection IE>presentation RPID>privacy=
15.4.10.2 Mapping SIP RPID header parameters to ISDN SETUP Information Elements
Three cases are illustrated to demonstrate the methodology used in translating the paramterters
Case 1 No RPID headers
SIP INVITE ISDN SETUP
Request-URI Called party number IE>number digits
User part of From: Calling party number IE>number digits
Name part of From: Display IE
Case 2 with calling RPID header
SIP INVITE ISDN SETUP
Request-URI Called party number IE >number digits
RPID>party=calling Calling party number IE
RPID>user= Calling party number IE>number digits
RPID>screen= Calling party number IE>screening indicator
RPID>privacy= Calling party number IE>presentation
RPID>display-name Display IE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 288 - 6/2/2009
Case 3 with calling and redirect RPID headers
SIP INVITE ISDN SETUP
Request-URI Called party number IE>number digits
RPID>party=calling Calling party number IE
RPID>user= Calling party number IE>number digits
RPID>screen= Calling party number IE>screening indicator
RPID>privacy= Calling party number IE>presentation
RPID>display-name Display IE
RPID>party=redirect Original called number / redirection IE
RPID>user= Original called number / redirection IE>number digits
RPID>screen= Original called number / redirection IE>screening indicator
RPID>privacy= Original called number / redirection IE>presentation
15.4.10.3 ISDN screening indicator to SIP screen Mappings
Screening indicator RPID>screen
User provided, not screened screen=no
User provided, verified and passed screen=yes
User provided, verified and failed screen=no
Network provided screen=no
15.4.10.4 SIP screen to ISDN screening indicator Mappings
RPID>screen Screening indicator
Screen=no User provided, not screened
Screen=yes User provided, verified and passed
15.4.10.5 Mappings between ISDN presentation indicator and SIP privacy
Presentation indicator RPID>privacy
Allowed privacy=off
Restricted privacy=on
15.4.11 RFC 3323 Pri vacy header and RFC 3325 extensions
SIP Vega gateways support the generation and reception of the Privacy header in INVITE and
REGISTER messages, as defined in RFC 3323, and also the P-Asserted-Identity and P-Preferred-
Identity headers defined in RFC3325.
The Privacy: header provides details about how the details relating to the calling party should be
handled.
To enable the generation and reception of the Privacy: header, set:
[_advanced.sip.privacy]
st andar d=r f c3323 ; def aul t =r f c3323
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 289 - 6/2/2009
The Privacy: header can include one or more of the following values:
header
11
session
12
user
none
id
13
optionally followed by
;critical
Note that if multiple types of privacy are required, all privacy types MUST be included in the
Privacy header field value.
header: Request that privacy services modify headers that cannot be set arbitrarily by the user
(Contact/Via). The user requests that those headers which might reveal information about the
user be obscured. Also, that no unnecessary headers should be added by the service that
might reveal personal information about the originator of the request.
session: Request that privacy services provide privacy for session media. The user requests that a
privacy service provide anonymisation for the session(s) initiated by this message. This will
mask the IP address from which the session traffic would ordinarily appear to originate. When
session privacy is requested, user agents MUST NOT encrypt SDP bodies in messages.
user: Request that privacy services provide a user-level privacy function. This privacy level is
usually set only by intermediaries, in order to communicate that user level privacy functions
must be provided by the network, presumably because the user agent is unable to provide
them. User agents MAY however set this privacy level for REGISTER requests, but SHOULD
NOT set 'user' level privacy for other requests. Any non-essential information headers are to be
removed and changes to From: and Call-ID: headers to make them anonymous is to be
performed.
none: Privacy services must not perform any privacy function. The user requests that a privacy
service apply no privacy functions to this message, regardless of any pre-provisioned profile for
the user or default behavior of the service. User agents can specify this option when they are
forced to route a message through a privacy service which will, if no Privacy header is present,
apply some privacy functions which the user does not desire for this message.
id: Privacy requsted for Third-Party Asserted Identity. The user requests that the Network
Asserted Identity to be kept private with respect to SIP entities outside the Trust Domain with
which the user is authenticated.
critical: Privacy service must perform the specified services or fail the request. The user asserts
that the privacy services requested for this message are critical, and that therefore, if these
privacy services cannot be provided by the network, this request should be rejected.
The extensions of RFC3325 add P-Asserted-Identity and P-Preferred_Identity.
P-Asserted-Identity: This is used between Trusted SIP entities; it carries the identity of the user
sending the SIP message as verified by authentication. There may be one or two
P-Asserted-Identity values. If there is one value, it MUST be a sip, sips, or tel URI. If there are
11
Not currently supported by the Vega
12
Not currently supported by the Vega
13
id is an extension to RFC3323 defined in RFC 3325
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 290 - 6/2/2009
two values, one value MUST be a sip or sips URI and the other MUST be a tel URI.
(Note: proxies can (and will) add and remove this header field.)
P-Preferred_Identity: This is used between a user agent and a Trusted Proxy; it carries the
identity that the user sending the SIP message wishes to be used as the P-Asserted-Header
that the Trusted Proxy will insert. There may be one or two P-Preferred-Identity values. If there
is one value, it MUST be a sip, sips, or tel URI. If there are two values, one value MUST be a
sip or sips URI and the other MUST be a tel URI.
(Note: proxies can (and will) add and remove this header field.)
15.4.11.1 ISDN to SIP
ISDN Presentation Indicator to SIP Privacy Header mapping:
ISDN Presentation Indicator SIP Privacy Header Content
Allowed Privacy: none
Restricted Privacy: id
Number not available Privacy: id
ISDN screening indicator to SIP P-Asserted-Identity / P-Preferred-Identity mapping
ISDN Screening Indicator SIP Header
Not screened P-Preferred-Identity
Passed P-Asserted-Identity
Failed P-Preferred-Identity
Network P-Asserted-Identity
e.g. Preferred Identity:
Pr i vacy: i d
P- Pr ef er r ed- I dent i t y: " St eve Hi ght " <si p: 4917@si p. vegast r eam. com>
e.g. Asserted Identity:
P- Asser t ed- I dent i t y: " St eve Hi ght " <si p: 4917@si p. vegast r eam. com>
P- Asser t ed- I dent i t y: t el : +441344784917
Pr i vacy: i d
15.4.11.2 SIP to ISDN
SIP Privacy Header to ISDN Presentation Indicator mapping:
SIP Privacy Header Content ISDN Presentation Indicator
Privacy: user Restricted
Privacy: none Allowed
Privacy: id Restricted
SIP P-Asserted-Identity / P-Preferred-Identity to ISDN screening indicator mapping
SIP Header ISDN Screening Indicator
P-Asserted-Identity Network
P-Preferred-Identity Not screened
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 291 - 6/2/2009
15.4.12 Session Timers
In order that SIP gateways can ensure calls are cleared down even if they never receive a BYE
message, session timers can be enabled. These are defined with the following parameters:
[sip]
sess_t i mer _i ndex=1
[sip.sess_timer.n]
enabl e=0
i nt er val =1800
mi n_i nt er val =300
r ef r esher _pr ef =r emot e
sess_t i mer _i ndex chooses the appropriate [sip.sess_timer.n] (n=1 to 3) set of
parameters to use. If enabl e=1 the Vega will act upon / generate session timer fields.
If the Vega initiates the SIP call it sends out an INVITE with the session timer value set to i nt er val ,
and the refresher parameter set to UAS or UAC depending on whether r ef r esher _pr ef is set to
remote or local (respectively). If r ef r esher _pr ef is set to local then the Vega will initiate the
session timer checks.
If a 422 response is received, the Vega will accept the higher requested session timer value.
If the Vega receives a call with the session timer value set, provided that the time is greater than
mi n_i nt er val then the Vega will accept the session timer value. It will accept the requested UAC /
UAS setting of the refresher parameter in the SIP message (initiating session timer checks if the
setting is UAS).
If the session time value received is smaller than mi n_i nt er val then the Vega will send out a 422
with the requested time set to mi n_i nt er val .
If the Vega is generating the session timer checks, after about half the negotiated session timer
timeout value (the session timer value both ends agree), the Vega will send out REINVITE
14
.
If it receives a 200 OK it re-starts the timer, if it does not receive the 200 OK after half the
time to the timeout it sends another REINVITE. If no 200 OK response is received by the
time the negotiated session timer timeout expires the call is cleared (a BYE is sent).
If the Vega is receiving the session timer checks, it too will count down the negotiated (agreed)
session timer timeout. If a REINVITE is received it will re-start the counter. If the countdown
expires then it will clear the call and send a BYE.
For more details on the Session Timers see RFC 4028.
14
Providing that there is enough time to do send out the REINVITE. To ensure the REINVITE is
sent, make sure that min_interval >=480ms.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 292 - 6/2/2009
15.4.13 Phone Context Headers
Phone contexts can be added to the To and From headers in SIP messaging for Vega initiated
calls using the table below, found on the SIP page of the web browser.
Local Phone-Contexts are used to populate the From header for ISDN to SIP calls based on the
values of TON (Type of Number) and NPI (Numbering Plan Information) in a received ISDN
SETUP message. They are also used to set the values for TON and NPI in the called party
number IE in the outgoing ISDN SETUP when a matching phone context is received in a SIP
INVITE.
Remote Phone-Contexts are used to populate the To header for ISDN to SIP calls based on the
incoming values of TON (Type of Number) and NPI (Numbering Plan Information) in an received
ISDN SETUP message. They are also used to set the values for TON (Type of Number) and NPI
(Numbering Plan Information) in the calling party number IE in the outgoing ISDN SETUP when a
matching phone context is received in a SIP INVITE
The following parameters have been added to configuration database for this feature (displayed
here with default values):
admin >show phone_context
[phone_context.local.1]
enable=1
[phone_context.local.1.pc.1]
NPI=any
TON=any
enable=0
name=local_phone.1.com
[phone_context.remote.1]
enable=1
[phone_context.remote.1.pc.1]
NPI=any
TON=any
enable=0
name= remote_phone.1.com
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 293 - 6/2/2009
Example SIP INVITE with phone-contexts setup:
SIP m:3809212 47750 00034--- UA TX --> To TCP(162):172.19.1.55:5060
INVITE sip:1234;phone-context=micrososft-ocs.com@default-reg-domain.com:5060;user=phone
SIP/2.0
Via: SIP/2.0/TCP 172.19.1.67:5060;branch=z9hG4bK-vega1-000A-0001-0004-CB9A50C9
From: "0" <sip:0;phone-context=microsoftremote-ocs.com@default-reg-domain.com>;tag=007D-
0006-DBDE6A28
To: <sip:1234;phone-context=micrososft-ocs.com@default-reg-domain.com>
Max-Forwards: 70
Call-ID: 0078-0004-63929283-0@91AD727D0597C801D
CSeq: 1523683 INVITE
Contact: <sip:0;phone-context=microsoftremote-ocs.com@172.19.1.67:5060;transport=tcp>
Supported: replaces, privacy
Allow: INVITE,ACK,BYE,CANCEL,INFO,NOTIFY,OPTIONS,REFER,UPDATE
Accept-Language: en
Content-Type: application/sdp
Privacy: none
P-Preferred-Identity: "0" <sip:0;phone-context=microsoftremote-ocs.com@default-reg-
domain.com>
User-Agent: VEGA400/10.02.08.2xS028
Content-Length: 294
v=0
o=Vega 134 134 IN IP4 172.19.1.67
s=Sip Call
c=IN IP4 172.19.1.67
t=0 0
m=audio 10008 RTP/AVP 18 0 8 4 101
a=rtpmap:18 G729/8000
a=rtpmap:0 PCMU/8000
a=rtpmap:8 PCMA/8000
a=rtpmap:4 G723/8000
a=rtpmap:101 telephone-event/8000
a=fmtp:101 0-15,16
a=fmtp:18 annexb=no
a=sendrecv
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 294 - 6/2/2009
15.4.14 User Defined String in SIP To / From Headers
User defined strings can be added to the SIP To and From and To headers sent by the Vega. A
typical use of this parameter is to add the user=phone parameter to SIP INVITEs sent by the
Vega.
Parameter:
_advanced. si p. f r om_header _ur i _par ams
Possible values:
NULL Def aul t Don t i ncl ude any st r i ng
Any st r i ng bet ween 1 and 39 char act er s i n l engt h
Parameter:
_advanced. si p. t o_header _ur i _par ams
Possible values:
NULL Def aul t Don t i ncl ude any st r i ng
Any st r i ng bet ween 1 and 39 char act er s i n l engt h
For example, if f r om_header _ur i _par ams=user =phone, a SIP INVITE would be similar to
this:
SI P m: 0626532 626532 00001- - - UA TX - - > To TCP( 72) : 172. 19. 1. 55: 5060
I NVI TE si p: 1234@def aul t - r eg- domai n. com: 5060; user =phone SI P/ 2. 0
Vi a: SI P/ 2. 0/ TCP 172. 19. 1. 67: 5060; br anch=z9hG4bK- vega1- 000A- 0001- 0000- 8C21B472
Fr om: " 0" <si p: 0@def aul t - r eg- domai n. com; user=phone>; t ag=007E- 0000- CB58C2DC
To: <si p: 1234@def aul t - r eg- domai n. com; user=phone>
Max- For war ds: 70
Cal l - I D: 0078- 0000- 61F25547- 0@91AD727D0597C801D
CSeq: 250611 I NVI TE
Cont act : <si p: 0@172. 19. 1. 67: 5060; t r anspor t =t cp>
Suppor t ed: r epl aces, pr i vacy
Al l ow: I NVI TE, ACK, BYE, CANCEL, I NFO, NOTI FY, OPTI ONS, REFER, UPDATE
Accept - Language: en
Cont ent - Type: appl i cat i on/ sdp
Cont ent - Lengt h: 294
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 295 - 6/2/2009
15.5 RFC2833
RFC2833 is a standard for transmitting and receiving DTMF signals and hookflash as part of the
real-time media stream.
For DTMF/hookflash to be sent as RFC2833 messages, firstly ensure that Out Of Band DTMF is
configured True against the appropriate codec.
15.5.1 RFC2833 Configuration
[sip]
dt mf _t r anspor t =r f c2833 ; use rfc2833 to send out-of-band DTMF (to use info
messages, set dtmf_transport=info; to transit both
RFC2833 and info messages, and to act upon received
RFC2833 messages, set dtmf_transport=rfc2833_txinfo)
r f c2833_payl oad=96 ; Configures the payload field in RTP messages for
RFC2833 data. RFC2833 data is sent in its own
UDP/IP packets (it is not combined with the audio).
[_advanced.rfc2833]
one_shot =0/ 1 ; In rfc2833 messages DTMF tone duration may or may
not be retained: 0 =true duration played, 1 =single fixed
length DTMF tone pulses played (on-time is defined by
_advanced.dsp.dtmf_cadence_on_time, off time defined
by _advanced.dsp.dtmf_cadence_off_time)
audi o_wi t h_DTMF=0/ 1 ; 0 =no audio packets are sent when RFC2833 tone
packets are sent; 1 =send both audio packets and
RFC2833 tone packets when tone present
t x_vol ume=0 t o 127 ; Power level of tone reported in Tx RFC2833 packets =-
n dBm0 (e.g. 10 =>-10dBm0). RFC2833 says tones
with a power 0 to -36dBm0 must be accepted, and
below -55dBm0 must be rejected.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 296 - 6/2/2009
15.6 Executive Interrupt
Vega gateways support Resource-Priority Headers for Preemption Events, as defined by RFC
4411 & RFC 4412.
This is a method by which calls from higher priority callers, rather than receiving a busy response
when making a call to a phone already engaged on a phone call, will bump the other party in the
conversation and will be connected directly to the called party. This feature is sometimes known
as Executive Intrusion, Boss / Secretary working, Call Barge, MLPP or Multi-Level Precedence
and Preemption.
If enabled, INVITES are sent out with Resource-Priority header values; also received INVITES
containing a Resource-Priorty header will not necessarily be rejected with busy, but will bump the
existing call if its Resource-Priority is higher than the Resource-Priority of the call in progress.
Call with precedence Y <=precedence X
A B C
INVITE
Resource-Priority: NameSpace.X
Etc.
INVITE
Resource-Priority: NameSpace.Y
Busy here
Call with precedence Y >precedence X
A B C
INVITE
Resource-Priority: NameSpace.X
Etc.
INVITE
Resource-Priority: NameSpace.Y
BYE
Reason: pre-emption ;cause=1
;text=UA Preemption
Etc.
If a call gets bumped the BYE for that call will contain a Reason header containing cause=1
;text=UA Preemption.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 297 - 6/2/2009
15.6.1 Configuring NameSpace for Resource-Priority Headers
A NameSpace defines a set of named priority values used in Resource-Priority headers. It is a
priority ordered list of priority names. Three standard NameSpace definitions are pre-configured in
the Vega: dsn, drsn and q735. Additional user defined NamesSpace definitions may be set up.
At any time the Vega only uses a single NameSpace definition to generate Resource-Priorities in
outgoing SIP calls and to act upon received Resource-Priorities in incoming SIP calls.
The NameSpace definition to use is configured in the Selected Namespace option.
If a call is received for a NameSpace other than that configured, the Vega will treat the call as
though it were a standard call with no Resource-Priority header.
Namespace definitions are priority ordered lists of names or IDs of priorities, listed in increasing
priority order.
e.g. dsn: lowest priority =routine
highest priority =flash-override.
Selecting modify in the user defined list allows the NamSpace Name and Priority values (IDs) to be
configured.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 298 - 6/2/2009
15.6.2 Resource-Priority for SIP calls initiated by Vega gateways
The Resource-Priority to use for outbound SIP calls is defined in the SIP authentication
configuration section.
A single Resource-Priority may be configured for each SIP Authentication User. (The subscriber
field defines which telephony port(s) the SIP Authentication User represents.)
The resource priority is configured through the selection of an entry in a pull down box. The values
contained in the pull down box are the values defined in the NameSpace configuration (see section
15.6.1 Configuring NameSpace for ).
The value selected will be the value sent out as the Resource-Priority with every SIP call made by
that user.
Ensure that the SIP Authentication User is enabled, otherwise
Resource-Priority handling will not be activated.
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 299 - 6/2/2009
15.7 SIP Music on Hold (MoH)
In default configuration, when a caller is put on hold they hear silence.
From Release 8.3 the Vega supports the playing of Music on Hold to the held party. Vega
gateways support the dr af t - i et f - si ppi ng- ser vi ce- exampl es- 11 method of supplying
music on hold.
This is easily configured through the web browser interface. On the SIP >SIP Music On Hold
Configuration page:
set up the SIP Music server
the URI is used to construct the SIP message
the IP / HostName and its IP port create the IP address to send the SIP messages to
Then select mode =sipping_service_11 to enable the dr af t - i et f - si ppi ng- ser vi ce-
exampl es- 11 method of supplying music on hold.
The dr af t - i et f - si ppi ng- ser vi ce- exampl es- 11 method operates as follows:
IP device being held Vega putting call on hold Music on hold server
Call in progress
INVITE (no sdp)
200 OK (sdp
m
)
(Re-)INVITE (sdp
m
)
200 OK (sdp
a
)
ACK
ACK (sdp
a
)
The Vega responds to 1xx provisional responses by opening media if an SDP body has been
received.
15.8 Multiple SIP Signalling Ports
Hookflash pressed to
put call on hold
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 300 - 6/2/2009
FXS gateways can optionally be configured to use a unique local SIP signalling port for each
configured SIP registration user. For instance, on a Vega 5000 24 port, provided each FXS port
has an associated registration user the Vega would use ports 5060 to 5083 for SIP signalling. This
can be particularly useful when working with a SIP proxy or softswitch that doesnt expect multiple
SIP UAs to be present behind a single IP address. i.e. Cisco Call Manager (CCM)
Parameter:
_advanced. si p. ci sco_cm_compat i bi l i t y
Possible Values:
0 Def aul t Do not use mul t i pl e SI P por t s
1 Use a di st i nct SI P por t f or each r egi st r at i on user
WARNING!
If this feature is enabled the local signalling port for TLS must
be set so that its outside the range that will be used for
multiple port signalling. The parameter that controls the TLS
port is sip.tls.local_rx_port.
15.9 TDM Channel Information
TDM (ISDN / POTS) B channel and interface information can be advertised in SIP messages using
'P-Access-Network-Info' headers.
In the case where a call originates from the Vega the header is included in the original SIP request
message (INVITE). In the case where the Vega terminates the call the header is included in the
ringing indication message (typically 180 or 183) or if not present in the 200 OK (connect)
message.
Parameter:
_advanced. si p. access_net wor k_i nf o. enabl e
Possible values:
0 Def aul t - Do not i ncl ude t he P- Access- Net wor k- I nf o header
1 - I ncl ude t he P- Access- Net wor k- I nf o header
Sample SIP message header:
SI P m: 0332867 18540 00124- - - UA TX - - > To UDP( 3) : 172. 19. 1. 58: 5060
I NVI TE si p: 123@def aul t - r eg- domai n. com: 5060 SI P/ 2. 0
Vi a: SI P/ 2. 0/ UDP 172. 19. 1. 81: 5060; r por t ; br anch=z9hG4bK- vega1- 000A- 0001- 0012
Fr om: " unknown" <si p: 17219158@def aul t - r eg- domai n. com>; t ag=007D- 0015
To: <si p: 123@def aul t - r eg- domai n. com>
Max- For war ds: 70
Cal l - I D: 0078- 000E- 66ACE409- 00000000@D02C806FC093603C6
CSeq: 133147 I NVI TE
P-Access-Network-Info: X-VEGA-NET;x-if0401;x-port0000;x-chan0001
Cont act : <si p: 17219158@172. 19. 1. 81: 5060>
Suppor t ed: r epl aces, pr i vacy
Al l ow: I NVI TE, ACK, BYE, CANCEL, I NFO, NOTI FY, OPTI ONS, REFER, UPDATE
Accept - Language: en
Cont ent - Type: appl i cat i on/ sdp
Cont ent - Lengt h: 294
In the example message header above the incoming ISDN call was placed using interface 0401 on
bearer channel 1.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 301 - 6/2/2009
15.10 SIP Status codes
15.10.1 1xx - SIP Provisional Responses Supported
The Vega responds to 1xx provisional responses by opening media if an SDP body has been
received.
1xx responses generated by the Vega are:
100 Trying - The Vega received an INVITE request and is processing it.
180 Ringing - The destination of the call is ringing.
181 Call is being forwarded
183 Session Progress - The call has not yet been answered but media is available.
Other 18x messages, like 182 Queued are accepted.
15.10.2 2xx - SIP Success Codes Supported
The Vega supports both 200 and 202 messages:
200 OK
202 Accepted - The Vega has accepted a transfer request and will generate an
INVITE to the transfer target.
15.10.3 3xx - SIP Redirection Codes Supported (Responded To)
The Vega responds to 3xx responses by trying to initiate another call if alternative "contacts" are
provided, otherwise the call is terminated.
300 Multiple Choices
301 Moved Permanently
302 Moved Temporarilly
305 Use Proxy
380 Alternative Service
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 302 - 6/2/2009
15.10.4 4xx - SIP Request Failure Codes Supported
With the exception of "401 Unauthorised", "407 Proxy Authentication Required", "415 Unsupported
Media Type" and 491 Request Pending, 4xx responses result in termination of the call.
4xx responses generated by the Vega are
15
:
400 Bad Request - Missing Call-ID field; the Vega received a request with a
"Call-ID" field that was missing or invalid.
400 Bad Request - Missing To field; the Vega received a request with a
"To" field that was missing or invalid.
400 Bad Request - Missing From field; the Vega received a request with a
"From" field that was missing or invalid.
401 Unauthorised (retry Register) - The Vega attempts to resend the INVITE with the
authentication response
[402 Payment Required]
[403 Forbidden]
404 Not Found - The Vega could not find a route for the destination
(sometimes caused by dial plan errors).
405 Method Not Allowed - The Vega received a request that it knows about but
does not allow. e.g. when a PRACK request is received
when sip.PRACK=off
406 Not Acceptable - The Vega received an INVITE with an illegal SDP.
407 Proxy Authentication Required - The Vega tries to resend the INVITE with the
authentication response
16
[408 Request Timeout - The server could not produce a response within a
suitable amount of time, for example, if it could not
determine the location of the user in time.]
409 Conflict
410 Gone
411 Length Required
413 Request Entity Too Large - the content length of a request must not exceed 1500
bytes.
414 Request-URI Too Long - The request-URI must not exceed 100 characters
415 Unsupported Media Type - The request received by the Vega has a message body
which is in an unsupported format. (Note: not
necessarily a media problem)
15
Items in square brackets are not generated by the Vega, but will be handled by the Vega.
16
408 is not generated by the Vega, but it will accept and handle it
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 303 - 6/2/2009
420 Bad Extension - The Vega did not understand the protocol extension
specified in a "Proxy-Require" or "Require" header.
422 Session Interval Too Small - The Session Interval requestsed is lower than the
min_interval configured in the Vega
480 Temporarily Unavailable - The Vega received a cause 18 (no user responding)
disconnection on its telephony interface.
481 Call Leg/Transaction Does Not Exist - The Vega received a request for which a
matching call leg and/or transaction was not found.
482 Loop Detected
483 Too Many Hops
484 Address Incomplete
485 Ambiguous
486 Busy Here - The destination of the call is busy.
487 Request Terminated - An INVITE request has been cancelled.
488 Not Acceptable Here - An INVITE was received for which no media is
supported. (i.e. expect Codec mismatch.) This will be
accompanied with a "304 No matching media" warning.
491 Request Pending - If the Call ID does not relate to this Vega, a REINVITE is
sent immediately. Otherwise, the Vega waits for the
other party to send a REINVITE
15.10.5 5xx - SIP Server Failure Codes Supported
The Vega responds to 5xx responses by terminating the call.
5xx responses generated by the Vega are:
500 Server Internal Error - No Call Legs Left; there are no more SIP resources
available
500 Server Internal Error - Still Processing Old Invite; an INVITE was received
while an earlier INVITE was still being processed.
500 Server Internal Error - Destination Out Of Order; the Vega received a cause 27
(destination out of order) on its telephony interface.
500 Server Internal Error - Temporary Failure; the Vega received a cause 41
(Temporary failure) on its telephony interface.
500 Server Internal Error - No Channel Available; the Vega received a cause 34
(no circuit/channel available) on its telephony interface.
500 Server Internal Error - Requested Channel Not Available; the Vega received a
cause 44 (Requested circuit/channel not available) on
its telephony interface.
501 Not Implemented - The Vega received a SIP request with a method it does
not recognise.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 304 - 6/2/2009
502 Bad Gateway
503 Service Unavailable - Includes Vega Congested.
504 Server Time-out
505 Version Not Supported - The Vega received a SIP request with a version other
than "SIP2.0".
513 Message Too Large
15.10.6 6xx - SIP Global Failure Codes Supported (Generated and Responded To)
The Vega responds to 6xx responses by terminating the call.
6xx responses generated by the Vega are:
600 Busy Everywhere
603 Decline - The Vega declined the request (in response to a REFER
request).
604 Does Not Exist Anywhere
606 Not Acceptable - If the Vega had previously sent a T.38 Fax INVITE, it will
try again with a G.711 INVITE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 305 - 6/2/2009
16 ENHANCED NETWORK PROXY
ENP (Enhanced Network Proxy) is a license enabled feature (i.e. requires a special license key to
be applied to the product). Please contact the supplier of your product to obtain an ENP license
key. ENP was previously referred to as VRP (Vega Resilient Proxy) in earlier firmware releases.
16.1 Description
The Enhanced Network Proxy feature (ENP) greatly extends the capabilities of a gateway product
by including SIP proxy functionality within a single device.
ENPs principle functions are twofold:
To provide resilience for local SIP UAs in case of loss of contact with ITSP proxy.i.e.
Through broadband failure, or loss of ITSP network connection.
To allow some calls that would normally always route to the ITSP to route to other
devices. These can include the local gateway (hosting ENP) or other gateways or SIP
devices.
16.2 ENPs Modes Of Operation
ENP can be configured to operate in three different modes (or disabled):
standalone_proxy
forward_to_itsp
itsp_trunking
off
Configuration via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Mode
Configuration parameter:
si ppr oxy. mode
16.2.1 Standalone Proxy Mode
In this mode the ENP behaves as a stand alone SIP Registrar and Proxy. The ENP can be used
for simple registration and proxy operations, enabling SIP devices to call one another, make (or
receive) calls via the gateway (for example to the PSTN or a PBX).
The ENP in standalone mode will support up to 120 attached (registered) endpoints (SIP devices).
The ENP supports basic call routing and SIP transfers, but does not provide more enhanced PBX
features such as Voice Mail.
Devices that wish to register to the ENP must either be defined as a SIP Proxy Auth User or have
an i.p. address defined in the Trusted IP Address table.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 306 - 6/2/2009
Devices defined as SIP Proxy Auth Users will be challenged for authentication, whereas devices
with i.p. addresses defined in the Trusted IP Address table will not be challenged for authentication
(they will just register).
Additionally Trunk Gateways (TGWs) can be defined. This enables calls to be routed to (and from)
TGWs without the need for the TGWs to register as endpoints. See further information regarding
TGWs below.
16.2.2 Forward To ITSP Mode
In this mode the ENP has one (or more) SIP ITSP Proxies defined in its configuration. All local (to
the ENP) SIP devices are configured to use the ENP as their outbound proxy. All SIP messaging is
sent via the ENP to the ITSP Proxy, and successful registrations are cached by the ENP.
Should the connection to the ISTP Proxy fail (the ENP continuously checks availability by sending
SIP OPTIONS messages) then all local devices with cached registrations will still be able to
communicate via the ENP. Once the ISTP Proxy connection is restored all SIP messaging is (once
again) sent via the ENP to the ITSP.
If a call is received and routing is configured such that a particular call is destined for a TGW then
the SIP messaging is forwarded to the TGW. See further information regarding TGWs below.
16.2.3 ITSP Trunking Mode
In this mode the ENP behaves similarly to the forward_to_itsp mode, however if a call is received
and is destined for a locally registered endpoint (Trusted IP Address, SIP Proxy Auth User or
TGW) then the SIP messaging will not be sent to the ISTP it will be routed directly to the local
endpoint destination (including TGWs).
16.3 ENP Configuration Details
16.3.1.1 ENPs Realm
The ENPs Realm should (in the case of working with an ITSP) be configured as the ITSP
realm/domain (i.e. myitsp.com). In the stand_alone mode the realm could be the i.p. address of
the gateway.
Configuration via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Realm
Configuration parameter:
si ppr oxy. r eal m
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 307 - 6/2/2009
16.3.1.2 ENPs Rx Port
The ENPs Rx (receive) Port should be different to the gateways Local SIP Port (configured in the
gateways SIP settings). It is useful to consider the ENP as a separate device to the gateway which
shares its i.p. address with the gateway, but is addressed using a different i.p. port.
When the gateway is sending SIP messaging to the ENP it can address it using the local loopback
address of 127.0.0.1 and the ENPs Rx Port.
Configuration via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Rx Port
Configuration parameter:
si ppr oxy. r x_por t
16.3.1.3 How Can I Tell Who Is Registered To The ENP?
All registered users (registered to the ENP, possibly to an ITSP too if set to forward_to_itsp
mode) can be seen in the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy Registered Users
The following CLI command will also show registered users:
si ppr oxy show r eg
16.3.1.4 SIP Proxy Auth Users
SIP Proxy Auth Users (as described above) are sip endpoints which are able to register directly
with the ENP. In forward_to_itsp mode endpoints do not necessarily need to be defined as
authentication users all registration requests are forwarded to the ISTP (if they are successful,
then the registration details will be cached in the ENP, ready to be used in the case of failure of the
ITSP link).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 308 - 6/2/2009
If the ENP is in forward_to_itsp or itsp_trunking modes then when the endpoint registers to the
ENP, the ENP will forward the registration messages to the ISTP. Should the ITSP reject the
registration then the endpoint will not be registered to the ENP (even if the SIP Proxy Auth User
information matches the endpoints registration request).
SIP Proxy Auth Users can be defined (and created) via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy Authentication Users
The following configuration parameters define a Sip Proxy Auth user (where x is the index of the
SIP Proxy Auth User, i.e. 1,2,3 etc.):
si ppr oxy. aut h. user . x. al i ases (see below)
si ppr oxy. aut h. user . x. enabl e (overall activation of SIP Proxy Auth User)
si ppr oxy. aut h. user . x. passwor d (SIP Proxy Auth User password same as ITSP if ITSP
is used)
si ppr oxy. aut h. user . x. user name ( SIP Proxy Auth User username same as ITSP if
ITSP is used)
16.3.1.5 SIP Proxy Auth User Aliases
Some ITSPs register using a different number from the PSTN number assigned to that device /
SIP user account. The ENP can support these user aliases, so (for example) in the event of an
ITSP failure other registered users can continue to call endpoints using the alias numbers.
Additionally the ENP can be configured to always use aliases to route calls to endpoints.
SIP Proxy Auth Users Alias control can be defined via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy Authentication Users > Use
Aliases
The following configuration parameter defines the ENP behaviour regarding aliases
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 309 - 6/2/2009
si ppr oxy. aut h. user . use_al i ases
16.3.1.6 SIP Proxy IP Filters
The following SIP Proxy IP Filters exist in the ENP:
Ignored IP Addresses
Rejected IP Addresses
Trusted IP Addresses
16.3.1.7 SIP Proxy IP Filters Ignored IP Addresses
SIP devices which signal the ENP using source i.p. addresses which are within a range defined as
ignored will not be responded to. This is to prevent SIP spamming where some device is
attempting to access (register to) the ENP to illegally gain access.
SIP Proxy Ignored IP Addresses can be defined via the Web User Interface (in ranges of i.p.
addresses):
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy IP Filters > Ignored IP
Addresses
The following configuration parameters define the SIP Proxy IP Filter Ignored IP Addresses
(where x is the index of the Ignored IP Address range, i.e. 1,2,3 etc.):
si ppr oxy. i gnor e. x. enabl e (overall control of ignored range index)
si ppr oxy. i gnor e. x. i pmax (i.p. range minimum value)
si ppr oxy. i gnor e. x. i pmi n (i.p. range maximum value)
16.3.1.8 SIP Proxy IP Filters Rejected IP Addresses
SIP devices which signal the ENP using source i.p. addresses which are within a range defined as
rejected will have their signalling requests actively rejected (with a SIP Forbidden response).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 310 - 6/2/2009
SIP Proxy Rejected IP Addresses can be defined via the Web User Interface (in ranges of i.p.
addresses):
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy IP Filters > Rejected IP
Addresses
The following configuration parameters define the SIP Proxy IP Filter Rejected IP Addresses
(where x is the index of the Rejected IP Address range, i.e. 1,2,3 etc.):
si ppr oxy. r ej ect . x. enabl e (overall control of reject range index)
si ppr oxy. r ej ect . x. i pmax (i.p. range minimum value)
si ppr oxy. r ej ect . x. i pmi n (i.p. range maximum value)
16.3.1.9 SIP Proxy IP Filters Trusted IP Addresses
By default, SIP devices with i.p. addresses which are not defined in any SIP Proxy IP Filter will
have their registration requests (in the case of fwd_to_itsp and itsp_trunking modes) forwarded to
the ISTP. It is up to the ITSP to challenge requests for authentication (which it may be configured
not to do).
If the link to the ITSP fails then the ENP will have responsibility for challenging requests for
authentication, so any devices which are not able to perform authentication functions will not be
able to process calls.
Defining a SIP devices i.p. address in the trusted i.p. address range allows these devices to
register to the ENP without any authentication challenges. If the ENP is in stand_alone mode a
trusted device will be allowed to register to the ENP without any challenges for authentication.
SIP Proxy trusted IP Addresses can be defined via the Web User Interface (in ranges of i.p.
addresses):
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP Proxy IP Filters > Trusted IP
Addresses
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 311 - 6/2/2009
The following configuration parameters define the SIP Proxy IP Filter Trusted IP Addresses
(where x is the index of the Trusted IP Address range, i.e. 1,2,3 etc.):
si ppr oxy. t r ust . x. enabl e (overall control of trusted range index)
si ppr oxy. t r ust . x. i pmax (i.p. range minimum value)
si ppr oxy. t r ust . x. i pmi n (i.p. range maximum value)
16.3.1.10 SIP ITSP Proxies
The ENP can be configured to use a single, or multiple, ISTP proxies when in forward_to_itsp or
itsp_trunking modes.
SIP Proxy SIP ITSP Proxies can be defined via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP ITSP Proxies
The following configuration parameters define the SIP Proxy SIP ITSP Proxies (where x is the
index of the SIP Proxy SIP ITSP Proxy, i.e. 1,2,3 etc.):
si ppr oxy. i t sp_pr oxy. x. enabl e (overall activation of ITSP connection usage)
si ppr oxy. i t sp_pr oxy. x. i pname (i.p. address or resovable name)
si ppr oxy. i t sp_pr oxy. x. por t (i.p. port to send SIP messages to ITSP proxy)
16.3.1.11 SIP ITSP Proxy Availability Test
By default the ENP checks for the availability of the ITSP proxy by sending SIP OPTIONS
messages to the remote platform(s) (every 30 seconds). BYE messages can also be used to poll
for availability this option is useful for those SIP devices that do not respond to OPTIONS
messages (e.g. Microsoft OCS).
If a response is not received the ITSP proxy is deemed down. If there are no available proxies
then the ENP behaves in failover mode, and allows locally registered endpoints to communicate
despite the unavailability of the ITSP proxy.
When in failover mode the ENP continues to test for ITSP proxy availability (by sending SIP
OPTIONS messages), should a response be received the ENP declares the ITSP as available (up)
and will once again route SIP messages to the ITSP.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 312 - 6/2/2009
SIP Proxy SIP ITSP Proxies Availability Test can be controlled (enabled or disabled) via the Web
User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP ITSP Proxies > Proxy Test
The following configuration parameter defines whether the SIP Proxy test is enabled:
si ppr oxy. i t sp_pr oxy. accessi bi l i t y_check
Note the signalling transport for the OPTIONS messages is also configurable (between UDP and
TCP) but only via the command line, using the following parameter:
si ppr oxy. i t sp_pr oxy. opt i ons_t r anspor t
Additionally, the following CLI command can be used to show the status of the remote proxies
(from the perspective of the ENP):
si ppr oxy st at us
16.3.1.12 Using Multiple SIP ISTP Proxies
When multiple ITSP proxies are defined they can be used in three different modes:
normal
cyclic
dnssrv
When set to normal mode (and if the SIP ISTP proxy is available) the ENP will use the first defined
SIP ITSP proxy. Should this primary SIP ITSP proxy become unavailable the ENP will use the next
available defined SIP ISTP proxy. Should there be no available SIP ITSP proxies the ENP will go
into failover mode.
When set to cyclic mode the ENP will use the defined available SIP ITSP proxies in a cyclic order
i.e. if there are three available proxies the ENP will use proxy 1, then proxy 2, then proxy 3 and
then proxy 1 again.
When set to dnssrv mode the ENP expects only a single SIP ITSP proxy to be defined in its
configuration. When the ENP tries to resolve the SIP ITSP proxy name the DNS server should
respond with available (multiple) proxy addresses with appropriate weighting for each. The ENP
sends OPTIONS messages to all the resolved SIP ITSP proxies to determine availability, and
respects the weighting set by the DNSSRV response for SIP traffic routing.
SIP Proxy SIP ITSP Proxies Mode can be configured via the Web User Interface:
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 313 - 6/2/2009
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP ITSP Proxies > Mode
The following configuration parameter defines what multiple SIP ITSP proxy mode is to be used:
si ppr oxy. i t sp_pr oxy. mode
16.3.1.13 SIP ITSP Proxies Signalling Transport
The signalling transport used for communication with the ITSP is configurable (between UDP and
TCP transports).
SIP Proxy SIP ITSP Proxies Transport can be configured via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > SIP ITSP Proxies > Transport
The following configuration parameter defines what SIP ITSP proxy transport is to be used:
si ppr oxy. i t sp_pr oxy. si g_t r anspor t
16.3.1.14 SIP Proxy Trunk Gateways
TGWs can be considered as SIP UAs (user agents) that can have calls routed to / from the ENP.
The principle difference between a TGW and registered endpoints is that TGWs routing is based
on routing rules defined in the ENP (where particular called numbers are routed towards the TGW),
not by virtue of being a registered endpoint.
TGWs can:
have availability checked using SIP OPTIONs messages (similar to ITSP Proxies).
be forced to authenticate with the ENP (similar to registered endpoints).
be utilised in a routing only, cyclic or weighted (dnssrv) modes.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 314 - 6/2/2009
TGWs are classified as either PSTN TGWs or Local TGWs. There are certain routing restrictions
applied to PSTN TGWs to prevent call looping in PSTN networks.
When a TGW is classified as a PSTN TGW the following routing restrictions apply:
calls from PSTN gateways cannot be routed to other PSTN gateways
calls from PSTN gateways cannot be routed to the ITSP
calls from unregistered users (even if " trusted") cannot be routed to PSTN gateways
By default the gateway hosting the ENP is considered as a PSTN TGW, and appears in the default
configuration (with the i.p. address of 127.0.0.1) as the first defined TGW. This first TGW definition
is not configurable.
SIP Proxy Trunk Gateways can be defined via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Trunk Gateways
The following configuration parameters are used to define a TGW (where x is the index of the
TGW, i.e. 2,3 etc. Note: t r unk_gw. 1 is not configurable):
si ppr oxy. t r unk_gw. x. enabl e (overall activation of TGW)
si ppr oxy. t r unk_gw. x. i pname (i.p. address or resoveable name)
si ppr oxy. t r unk_gw. x. i s_pst n_gw (flags if TGW is defined as a PSTN TGW)
si ppr oxy. t r unk_gw. x. por t (i.p. receive port of the TGW)
Further Trunk Gateway configuration can be defined via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Trunk Gateways
The following configuration parameters are used to define additional TGW controls:
si ppr oxy. t r unk_gw. accessi bi l i t y_check
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 315 - 6/2/2009
(controls use of SIP OPTIONS or BYE messages to check TGW availability)
si ppr oxy. t r unk_gw. al l ow_i t sp_cal l s_t o_pst n
(controls ability for ITSP calls to be routed to PSTN TGWs)
si ppr oxy. t r unk_gw. f r om_act i on
(controls whether TGWs are trusted (do not register), required to authenticate, actively rejected or
ignored)
si ppr oxy. t r unk_gw. mode
(controls mode in which TGWs can be load shared or not)
si ppr oxy. t r unk_gw. opt i ons_t r anspor t
(controls signalling transport for SIP OPTIONs messages)
si ppr oxy. t r unk_gw. si g_t r anspor t
(controls signalling transport for TGW SIP messages)
There are two additional routing restriction configuration parameters available which control
routing towards the ITSP when the ENP is configured in forward_to_itsp mode.
si ppr oxy. t r unk_gw. f or war d_t o_i t sp_mode. al l ow_l ocal _t r unk_cal l s_t o_i t sp
si ppr oxy. t r unk_gw. f or war d_t o_i t sp_mode. al l ow_pst n_cal l s_t o_i t sp
by default both of these parameters are set to never.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 316 - 6/2/2009
16.3.1.15 Trunk Gateway Call Routing
Routing of calls towards the TGWs is defined as a series of routing plans, where call routing
decisions can be made based on the following call attributes:
TEL: (called number)
TELC: (calling number)
TAC: (calling i.p. address)
If a call is received that matches a routing plan (i.e. the called number matches the TEL: call
attribute in a routing rule) then the call is routed to a defined TGW (or to a single TGW from a
defined list of TGWs).
Where a list of multiple TGWs is defined in a routing rule (in a comma separated list), the choice of
which TGW to use can be defined as:
linear_up (i.e. the first TGW defined in the list of TGWs is routed to first if the call fails or
the TGW is unavailable the second defined TGW is used etc.)
equal (i.e. all defined TGWs are routed to equally pseudo randomly)
weighted (i.e. 60:40 for two defined TGWs)
The range of SIP error responses which trigger a re-attempt to the next available TGW can be
defined (by default 500-599 responses will trigger the ENP to attempt a call to the next available
TGW).
Trunk Gateway Call routing can be configured via the Web User Interface:
Expert Config > SIP Proxy > SIP Proxy Configuration > Trunk Gateway Call Routing
The following configuration parameters are used to define the Trunk Gateway Call Routing (where
x is the routing plan index, i.e. routing rule 1,2,3):
si ppr oxy. t r unk_gw. pl an. x. dest
(Call attributes, if matched use this routing plan)
si ppr oxy. t r unk_gw. pl an. x. enabl e
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 317 - 6/2/2009
(Overall activation of the routing plan)
si ppr oxy. t r unk_gw. pl an. x. gw_l i st
(Comma separated list of TGW ids)
si ppr oxy. t r unk_gw. pl an. x. name
(A name assigned to the plan)
si ppr oxy. t r unk_gw. pl an. x. r edi r ect i on_r esponses
(The range of error responses on which to attempt the call to the next TGW)
si ppr oxy. t r unk_gw. pl an. x. r out i ng_r ul e
(TGW routing rule i.e. linear_up, equal or weighted i.e. 20:30:50)
16.3.1.16 PSTN Gateway Fallback
In stand_alone mode if a call is received from a TGW or a registered endpoint and the called
number is not a registered endpoint and there is no matching TGW routing, the call will be routed
out to the PSTN Fallback Gateway.
In forward_to_istp or itsp_trunking modes if a call is received from a local TGW or a registered
endpoint and the called number is not a registered endpoint and there is no matching TGW routing,
the call will be routed out to the PSTN Fallback Gateway.
The PSTN Fallback Gateway can be defined as all gateways defined in the TGW list or a list of
specified TGW identifiers (with the same routing decision rules as in the TGW routing i.e.
linear_up, equal or weighted (i.e. 20:80).
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 318 - 6/2/2009
The range of SIP error responses which trigger a re-attempt to the next available PSTN TGW can
be defined (by default 500-599 responses will trigger the ENP to attempt a call to the next available
PSTN TGW).
16.3.1.17 Checking If Unit Has SIP PROXY License
ENP is a licensable feature, in other words a special license key must be applied to the gateway to
enable the ENP feature to be used.
To check if the gateway has the appropriate license key from the CLI type:
upgr ade
l i cense
In the output ensure that the active license key confirms that the SIP PROXY feature is available:
syst eml i censed f or SI P PROXY
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 319 - 6/2/2009
17 SNMP MANAGEMENT
Vega gateways contain an SNMP server that is compatible with SNMP versions 1 and 3,
supporting MIB-1 and MIB-2 definitions. The Vega will also generate SNMP traps on key system
events.
17.1 SNMP Configuration
To enable SNMP the following information will need to be configured:
[snmp.mib2.system] ; basi c SNMP syst emdet ai l s
sysCont act ; cont act name f or t hi s Vega
sysLocat i on ; l ocat i on det ai l s f or t hi s Vega
[snmp.mib2.managers.n] ; def i ni t i on of who i s al l owed t o manage t he Vega
i p ; manager s i p addr ess
subnet ; mask t o i dent i f y si gni f i cant par t of manager s I P
; addr ess t o check
communi t y ; communi t y name ( one of t he mi b2. communi t i es. m. name)
suppor t _snmpv3 ; Enabl e / di sabl e SNMP V3 suppor t ( di sabl ed = v1)
[snmp.mib2.communities.m] ; l i st of avai l abl e communi t i e
name ; communi t y name
get ; get al l owed ( 1=yes, 0=no)
set ; set al l owed ( 1=yes, 0=no)
t r aps ; t r aps al l owed ( 1=yes, 0=no)
A list of allowed managers must be configured as only members of this closed user group are
allowed access to the SNMP variables and can receive SNMP traps. The contact and location
details can be altered using the corresponding SNMP set commands via a manager.
17.2 SNMP Enterprise Object-ID
The VegaStream Object-ID for Vega gateways is: 1.3.6.1.4.1.4686.11
1 (ISO).3 (organisations).6 (dod).1 (IAB Administered).4 (private).1 (enterprises).4686 (enterprise ID - VegaStream).11 (Vega)
17.3 Trap Support
Support is available for the following traps:
Trap Number Definition
0 System Cold Boot
1 System Warm Boot
2 Link Down
3 Link Up
4 Authentication Failure
6 Enterprise specific see specific codes for details
For details of the enterprise specific trap specific codes and for further details on SNMP,
see Information Note IN 08 SNMP management
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 320 - 6/2/2009
18 UPGRADES AND MAINTENANCE
18.1 Upgrading the Vega Firmware
Upgrading firmware is a relatively easy task. Full upgrade instructions are provided along with the
firmware file itself, normally in the same zipped folder. See www.vegaassist.com for further
information on how to obtain new firmware.
18.2 The Boot-time Recovery Menu
Vega Boot code supports a couple of disaster recovery functions to assist the user in extreme
circumstances.
NOTE
Use of these functions can seriously affect the configuration of your
Vega - Only use these functions under the direction of your supplier
To access the boot menu you will need the following:
Straight through DB9 to RJ 45 RS-232 serial cable
Terminal DTE or PC based terminal emulator application (like Microsoft Hyper Terminal)
configured for 9600 bps, 8/N/1
Power the Vega off and then on, and in the first 10 seconds press and hold the enter key on the
terminal/emulator application keyboard. A message will appear saying Press Y for boot menu.
At this point press the Y key, and a menu will appear with the following options:
Reset Syst emConf i gur at i on and Cl ear Passwor ds
Swi t ch Act i ve Boot Par t i t i on
Exi t boot menu
18.2.1 Reset System configuration and Clear Passwords
Select Reset Syst emConf i gur at i on and Cl ear Passwor ds from the menu, and press
Y to confirm your choice. The configuration and passwords on the Vega will be reset back to
factory defaults.
WARNING!
Unlike the FACTORY RESET command, this BOOT MENU
operation will erase ALL data in the Vega, and restore ALL
settings back to factory default values (including, for example,
lan.if.x.ip and all passwords). Any license applied will also be
removed. This could result in severe loss of service.
18.2.2 Switch Acti ve Boot Partition (- Reverting to a Previous Firmware Image)
Select Swi t ch Act i ve Boot Par t i t i on from the menu. A list of up to two runtime images
will then be displayed, labelled 1. and 2., with their corresponding firmware version and build
details. The current partition will be displayed as CURRENT. To switch to the other runtime
partition select the appropriate number and then confirm your choice.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 321 - 6/2/2009
There will be a pause while partitions are swapped and then the Vega will automatically re-boot in
order to start running from this partition.
NOTE
You should carry out a factory reset after a change in firmware
partition to ensure that all parameters are appropriately initialised
for this version of code.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 322 - 6/2/2009
19 AUTOEXEC SCRIPT
The autoexec script function allows a Vega to automatically upgrade its firmware and configuration
on power up and re-boot. The contents of the autoexec script file defines the exact operations that
the Vega will make.
This script is downloaded as a file from a server (e.g. tftp. ftp, http or https) and executed. The
collection and execution of autoexec files is triggered by:
Power on
Reboot
Scheduled autoexec
SIP Notify
Trying to collect and execute an autoexec file at power on and Vega reboot is enabled by default;
scheduled autoexec needs to be configured.
The method for collecting the autoexec file (tftp, ftp, http, https) will be dependent on the setting of
l an. f i l e_t r ansf er _met hod. If it succeeds it will then execute the commands within that script
file.
Also see the document IN_42-Vega_Provisioning available on www.VegaAssist.com
19.1 The Script File
The script file contains a set of CLI commands that are executed on boot-up.
While the script file can run most CLI commands, the script file typically contains:
1) A CLI command to download a specific firmware.
2) A CLI command to load a specific configuration.
3) Optionally, a few CLI commands to set some specific config parameters.
The script file is not intended to contain more than a few lines of configuration data and must be
less than 512 bytes.
19.2 A Typical Script File
upgr ade
downl oad enabl e
downl oad f i r mwar e vega50pwi sc. abs r eboot i f new
exi t
get conf i g2. t xt save r eboot i f di f f
This script file will make sure that the Vega will load the vega50pwi sc. abs firmware and the
conf i g2. t xt configuration file.
NOTE
There MUST be a blank line after the last command line in the
autoexec script file as the Vega needs to see the Carriage Return at
the end of the command line in order to execute the command.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 323 - 6/2/2009
19.3 Script File - Permitted Command Set
For security reasons, the command set for the script file is a subset of the full Vega command set,
for instance it is not possible to change the password from the script file. Commands that are
supported include:
APPLY
BI LL [ OFF| ON| Z| CLEAR]
BI LL DI SPLAY [ OFF| ON]
BLOCK CALLS
BOOT MANAGER
CD
CLEAR STATS
CP
DELETE
DOWNLOAD ENABLE
DOWNLOAD BOOT
DOWNLOAD FI RMWARE
GET
NEW
ON ERROR BLOCK
ON ERROR RUN
PART1
PART2
PURGE
PUT
SAVE
SET
SHOWBANNER
SHOWBI LL
SHOWCALLS
SHOWHOSTS
SHOWPORTS
SHOWSTATS
SHOWVERSI ON
TGET
TPUT
UNBLOCK CALLS
UPGRADE
19.4 CLI Command Extensions
In order to allow commands to be processed conditionally, a number of extensions to existing
commands have been implemnented:
( 1) get conf i g. t xt i f di f f
Same as get but before loading the configuration the Vega checks the version of the new
configuration file against that specified at _advanced.autoexec.lastconfig. The configuration file is
only loaded if the version is different.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 324 - 6/2/2009
In a file that has been created using the Vega's put or "t put " command, the configuration
version is identified by the VEGACONFIGVERSION header at the head of the file:
;
; Scr i pt gener at ed usi ng
; PUT hel . t xt <al l >
; VEGACONFI GVERSI ON: Vega50WI SC: 01/ 01/ 1999 00: 03: 00
;
Therefore, if the ifdiff parameter is specified, if _advanced.autoexec.lastconfig is
"Vega50WISC:01/01/1999 00:03:00", then the config will not be loaded.
( 2) get conf i g. t xt save r eboot i f di f f
Same as the "get conf i g. t xt i f di f f " except that if the get is performed the Vega will save
the config and then reboot.
( 3) get conf i g. t xt save r eboot i f needed i f di f f appl y
Same as the "get conf i g. t xt save r eboot i f di f f " except that the reboot will only occur
if there are config variables that have changed that need the Vega to be rebooted to activate them.
apply is necessary to apply parameters if the reboot is not needed.
( 4) get conf i g. t xt save r eboot i f neededwheni dl e i f di f f appl y
Same as the "get conf i g. t xt save r eboot i f needed i f di f f " except that if the reboot is
needed it will be delayed until there are no calls in progress on the Vega.
( 5) downl oad f i r mwar e vega50pwi sc. abs i f new
Same as "download firmware" but before loading the code the Vega checks the version of code on
the sever against the current version. The firmware will only be loaded if the code on the server is
newer.
The current version is shown when you do "show ver si on":
e. g.
Ver si on: 04. 02. 04
Bui l t : May 9 2001 14: 42: 14 T001
In a version description there is:
Ver si on: <HW>. <SWmaj >. <SWmi n>
Bui l t : <Dat e> <Ti me> T<Bui l dTag>
The <Date>and <Time>fields are not checked but the other fields (in order of importance, most
important first) are :
<HW> - har dwar e ver si on
<SWmaj > - f i r mwar e maj or ver si on
<SWmi n> - f i r mwar e mi nor ver si on
<Bui l dTag> - t ag I D whi ch t oget her wi t h <HW>, <Swmaj > and <Swmi n> make
t hi s bui l d I D uni que
Format of fields (lowest value first):
<HW> - 01, 02, 03, et c.
<SWmaj > - 01, 02, 03, et c.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 325 - 6/2/2009
<SWmi n> - 01, 02, 03, et c.
<Bui l dTag> - 001, 002, 003, et c.
If the "i f new" directive is specified, the following checks are performed in the following order:
( 6) downl oad f i r mwar e vega50pwi sc. abs i f di f f
Same as "download firmware" but before loading the code the Vega checks the version of the code
on the server against the current version. The firmware will only be loaded if the code version on
the server is different.
( 7) downl oad f i r mwar e vega50pwi sc. abs r eboot i f new
Same as "download firmware vega50pwisc.abs ifnew" except that if the download is performed the
Vega will automatically reboot.
( 8) downl oad f i r mwar e vega50pwi sc. abs r eboot i f di f f
Same as "download firmware vega50pwisc.abs ifdiff" except that if the download is performed the
Vega will automatically reboot.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 326 - 6/2/2009
19.5 Configuring Autoexec Parameters
The default configuration is:
[_advanced.autoexec]
enabl e=1
l ast conf i g=none
scr i pt f i l e1=%i scr i pt . t xt
scr i pt f i l e2=def aul t scr i pt . t xt
Term Description
enabl e The Vega will only try to fetch a script file if this is set to '1'.
l ast conf i g The version of the last successfully loaded configuration file this is updated by the
vega based on the last configuration loaded; there is no need to alter this parameter.
scr i pt f i l e1 The first file containing the commands to be executed on boot up.
scr i pt f i l e2 If the Vega can't find scriptfile1 then it will try scriptfile2.
19.6 Scriptfile Name Expandable Characters
In "_advanced. aut oexec. scr i pt f i l e1" and "_advanced. aut oexec. scr i pt f i l e2", the
expandable characters %i and %n can be used:
%i
Expands to the ip_address of the Vega. So, if the Vega's IP address is
aaa.bbb.ccc.ddd then "%i" will become "aaa_bbb_ccc_ddd". The IP address is taken
either from "lan.if.1.ip" in the configuration or from that obtained via DHCP (for Lan
interface 1).
%m
Expands to the MAC address of the Vega.
%n
Expands to the hostname of the Vega. The hostname is specified by "lan.name" in
the configuration.
%p
Expands to the product type as shown in show banner, e.g. VEGA400 and VEGA-6x4
e.g. if
[_advanced.autoexec]
scr i pt f i l e1=vega_%i _cf g. t xt
and the ip address of the vega is 192.168.1.102, then autoexec will look for a file
vega_192_168_1_102_cfg.txt on the tftp or ftp server.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 327 - 6/2/2009
19.7 Status Reporting
To report the success or failure of the firmware and configuration parameter loading, Vegas use
Alert log messages and SNMP "enterprise-specific" traps. The traps show up as:
trap objectID=enterprises.4686.11 and
trap specific code=x,
where x is the specific code for the enterprise trap (see Information Note IN-08 SNMP
management for values).
For example, on the CastleRock SNMP manager enterprise traps are displayed in the form:
ent er pr i ses. 4686. 11. 6. x
19.8 Example Sequence of Events
For the following script file:
upgr ade
downl oad enabl e
downl oad f i r mwar e vega50pwi sc. abs r eboot i f new
exi t
get conf i g2. t xt save r eboot i f di f f
The full sequence of events of an error-free execution of the above script is:
1) The Vega will fetch the script filefrom the ftp or tftp server
2) The Vega will download the new firmware if it is newer than the current version.
** VEGA WILL REBOOT **
3) The Vega will fetch the script file again.
4) It won't download the firmware because the firmware is already up-to-date (server version
of firmware is no longer newer).
5) It will load the config file config2.txt if it is different to the current loaded version.
6) The config will be saved.
** VEGA WILL REBOOT **
7) The script file will be fetched again.
8) The vega won't do the firmware download.
9) The vega won't do the config load.
10) The vega starts normal operation.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 328 - 6/2/2009
Once step 10 has been reached, if the Vega is rebooted again, the traps sent out by the Vega will
be:
ent er pr i ses. 4686. 1. 6. 22 firmware not loaded because it isn't new
ent er pr i ses. 4686. 1. 6. 21 config not loaded because the version isn't different
19.9 SIP Notify triggered autoexec
Using a SIP notify, the Vega can be requested to download and execute an autoexec file. The
structure ofg the autoexec is:
SIP m:1480342 141002 00009<-- UA RX --- From UDP(18):172.19.1.233:5060
NOTIFY sip:service@172.19.1.230:5060 SIP/2.0
Via: SIP/2.0/UDP 192.168.0.1:5060;branch=z9hG4bK-14823-1-0
From: sipp <sip:sipp@192.168.0.1:5060>;tag=14823SIPpTag001
To: sut <sip:service@172.19.1.230:5060>
Call-ID: 1-14823@192.168.0.1
CSeq: 1 NOTIFY
Contact: sip:sipp@192.168.0.1:5060
User-Agent: Provisioning
Event: ua-profile
Max-Forwards: 70
MIME-Version: 1.0
Content-Type: message/external-body; access-type="URL";
URL="http:/Steve/VegaStream/005058040070_notify.txt";
Content-Length: 0
This requests the Vega to download and execute the autoexec file
/Steve/VegaStream/005058040070_notify.txt from an http server.
When the Notify is received, the Vega will ask for authentication to ensure that only authorised
requests may cause the Vega to download new configuration.
For details on how to configure SIP Notify handling, see the document IN_42-Vega_Provisioning
available on www.VegaAssist.com
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 329 - 6/2/2009
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 330 - 6/2/2009
20 WORKING WITH FIREWALLS
The main job of a firewall is to block LAN traffic that is not known to be acceptable. One of the major
problems that VoIP introduces to firewall protection is the number of IP port numbers that the protocol
specifies as valid for carrying the media. Unless the Firewall is VoIP aware and can open and close IP
port numbers based on the protocol messages, the port number range that needs to be left open (i.e.
unprotected) is that specified by the RTP spec, 10,000 to 20,000.
In order to reduce the size of the hole that must be opened in the firewall, the Vega can be configured to
use a more limited subset of IP port numbers for receiving RTP media traffic. When it specifies the IP
port number for the far end device to send the media to, it looks in its configuration parameters for the
range of values it has been configured to use. By default the range 10,000 to 20,000 is configured (as
per the RTP specification).
If a lesser range is required, the Vega can be configured with up to 10 blocks of port numbers, allowing
islands of non-intersecting port numbers to be used for the media.
For example if the ranges 10,000 to 10,249 and 11,000 to 11,249 are to be used for media, then
configure the Vega as follows:
[_advanced.lan.port_range.1]
max=10249
mi n=10000
name=r t p_r ange1
pr ot ocol =udp
[_advanced.lan.port_range.6] / / used 6 as 2. . 5 ar e def i ned by def aul t
max=11249
mi n=11000
name=r t p_r ange2
pr ot ocol =udp
[_advanced.lan.port_range_list.1]
l i st =1, 6 / / _advanced. l an. por t _r anges 1 & 6 = r t p por t s
name=r t p_por t s
[_advanced.media]
r t p_por t _r ange_l i st =1 / / rtp port list defined by _advanced. l an. por t _r ange_l i st . 1
NOTE
The defined range must allow room for both RTP connections and
RTCP connections. By definition an RTP port is an even
numbered port and the associated RTCP port is the next higher
odd numbered port. To avoid problems of lack of RTP/RTCP
ports for media, the minimum number of ports that must be
supported over all the f i r st / l ast blocks must be 2 * Vega
ports.
To ensure that each RTP port can be used (because there is an
associated valid RTCP port) always make f i r st an even number
and l ast an odd number.
20.1 NAT
NAT Network address translation, is typically used to hide a network of private IP addresses
behind one or more public IP addresses. A NAT device changes the IP address and often the IP
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 331 - 6/2/2009
port number of the IP messages as they cross it. This causes problems to VoIP systems as the
VoIP protocol contains references to explicit IP addresses and port numbers, which typically do not
get translated.
Vega gateays have configuration parameters that allow it to operate with statically configured NAT
devices. This functionality allows the Vega to pre-change the in-protocol IP address and port
number information, so that they are consistent with the changes that the NAT device will make to
the message headers.
For further details on the problems of NAT, and for details on how to configure the Vega to
work with statically configured NAT devices, see information note IN 14 NAT handling
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 332 - 6/2/2009
21 QUALITY OF SERVICE (QOS)
Quality of Service is a whole network requirement. All switches / routers and other devices in the
LAN path as well as the endpoints must support and be configured correctly to support QOS,
otherwise any point that does not properly support QOS will be the weak link that loses or delays
packets and ruins the quality for the whole system.
It is up to end-points like Vegas to mark LAN packets appropriately so that the in-network
routers can give them the priority over other less time critical data transfers.
Vegas support QOS marking of LAN packets. They also support the generation of QOS reports
and the monitoring and logging of QOS events.
21.1 QOS marking of LAN packets
Vega units support the configuration of both i) Type of Service/Diffserv field in the IP header, and
ii) 802.1p/q fields in the Ethernet header.
WARNING!
802.1 Ethernet packet headers are 4 bytes larger than standard
Ethernet headers, and so use of 802.1p/q may not be backward
compatible with existing Ethernet systems only enable 802.1
p/q functionality on your Vega if your network supports these
LAN packets, otherwise you may lose LAN connection with it.
21.1.1 Layer 3 (IP header) Type Of Service bits
Vegas support the configuration of Internet Protocol Header Type Of Service (TOS) value. This is
a layer 3 value that LAN routers and switches can use to determine the priority of the IP packet in
comparison to other suitably tagged packets.
Configuration of Type Of Service parameters is performed using QOS profiles defined below in
section 21.1.3.3 QOS profile configuration.
The way the Type Of Service bits are used depends on the network manager. The original
specification of the TOS bits defines a general structure for using the bits. DiffServ refines and
makes more specific the use of the values. The use of the TOS bits in various scenarios is defined
below, however a fuller discussion may be found at:
http://www.aarnet.edu.au/engineering/networkdesign/qos/precedence.html
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 333 - 6/2/2009
21.1.1.1 Type Of Service values
The Type Of Service octet contains a 3 bit precedence value and 4 bits used to request minimize
delay, maximize throughput, maximize reliability, and minimize monetary cost the least
significant bit of the octet must remain zero.
In RFC1349 the Type Of Service value is defined as:
MS 3 bits =Precedence
Next 4 bits =Type Of Service
LS bit =Zero
The 3 bit Precedence field gives an increasing set of precedence:
000 -- priority 0, normal precedence
to
111 -- priority 7, network control (maximum precedence)
The value of Precedence used will depend on the design of the Network (and configuration of the
Network routers), but in typical networks a good value for precedence for VoIP traffic is 5.
The 4 bit TOS field is constructed from the following bitmaps:
1000 -- minimize delay
0100 -- maximize throughput
0010 -- maximize reliability
0001 -- minimize monetary cost
0000 -- normal service
21.1.1.2 Diffserv
Diffserv is a specification that formalises the use of the TOS octet. From RFC2597, Diffserv has a
notion of two data transfer schemes, AF Assured Forwarding and EF Expedited Forwarding
In Assured Forwarding, at LAN routers / switches:
short term congestion will result in packets being queued
long term congestion results in packets being dropped
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 334 - 6/2/2009
Assured Forwarding uses 6 bits to identify 4 classes and 3 drop precedences (the 2 LS bits of the
TOS octet remain zero):
Class 1 Class 2 Class 3 Class 4
Low Drop
precedence
AF11 =00101000
(=40, 0x28)
AF21 =01001000
(=72, 0x48)
AF31 =01101000
(=104, 0x68)
AF41 =10001000
(=136, 0x88)
Medium Drop
precedence
AF12 =00110000
(=48, 0x30)
AF22 =01010000
(=80, 0x50)
AF32 =01110000
(=112, 0x70)
AF42 =10010000
(=144, 0x90)
High Drop
precedence
AF13 =00111000
(=56, 0x38)
AF23 =01011000
(=88, 0x58)
AF33 =01111000
(=120, 0x78)
AF43 =10011000
(=152, 0x98)
Expedited Forwarding implies that this traffic is high priority traffic and should take precedence
over ALL other LAN traffic. Packets are marked EF when they need to be transmitted across the
Network with low latency and low jitter.
In Expedited Forwarding:
This traffic takes precedence over all other traffic so long as the traffic rate stays within
preset bounds.
If the traffic rate is exceeded then the excess packets are dropped
Expedited Forwarding uses a single 6 bit value for identification (RFC2598), the 2 LS bits remain
zero:
10111000 (=184, 0xb8)
For VoIP traffic it is recommended that Expedited Forwarding is selected (set the TOS value to 184
(0xb8)).
21.1.2 Layer 2 (Ethernet Header) 802.1p Class of Service tagging and 802.1q VLAN tagging
Vegas support the configuration of both 802.1p Class of Service tagging and 802.1q VLAN tagging.
802.1 p/q are layer 2 (Ethernet header) values that LAN bridges, layer 2 routers and switches can
use to determine the priority of the IP packet in comparison to other suitably tagged packets.
WARNING!
802.1 Ethernet headers are 4 bytes larger than standard
Ethernet headers, and so may not be backward compatible
with existing Ethernet systems only enable 802.1 p/q
functionality on your Vega if your network supports these
packets, otherwise you may lose LAN connection with it.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 335 - 6/2/2009
If the Vega gateway is connected to an access port of an 802.1 p/q
switch/router, you do not need to enable 802.1 p/q handling on the
Vega because the switch/router will handle (add) the 802.1 p/q
labelling of the LAN packets.
Only enable 802.1 p/q handling on the Vega if you need the Vega
to specify the CoS (Class of Service / User Priority) or VLAN
membership, or if you want to connect the Vega to a trunk port of
an 802.1q enabled switch/router.
(A switch/router access port generally accepts both tagged and
untagged LAN packets the untagged packets will be assigned a
VLAN ID and priority by the switch/router. VLAN tagged packets
will usually be rejected if the VLAN ID is not the same as that
configured for this port.
A trunk port will generally accept only VLAN tagged LAN packets
it will not check the VLAN ID it will just pass on all packets)
The 802.1p (priority) can take a value in the range 0..7
0 =best effort priority really depends on configuration of network bridges, layer 2 routers and
switches
1 to 7 =increasing priority; 7=highest priority
The 802.1q (Virtual LAN) defines a LAN ID which can take a value in the range 0 to 4095
21.1.3 Configuring QOS Profiles
For flexibility Vegas support the ability to configure a number of QOS profiles. The QOS profile
that is used on a specific LAN packet depends on the currently active QOS profile. The active
QOS profile is specified using configuration parameters in the Vega. If the LAN packet relates to a
specific call, the dial planner can override the selection of QOS profile to be used.
The QOS profile to use is specified within a LAN_profile. The various LAN applications call up
which LAN profile (and therefore which QOS profile) to use for that appluication (e.g. calls, tftp, ftp
etc.).
21.1.3.1 Configuring QOS Profiles
The Qos profile to use in a specific circumstance is now selected by the LAN profile that has been
selected for that circumstance. LAN profiles enable both the selection of a physical LAN interface
(important for Vega 400) and the qos profile to use on that interface.
LAN profiles are defined for:
ftp
h.323
h.323 gatekeeper
http
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 336 - 6/2/2009
lan
ntp
sip
telnet
tftp
The Vega will use the qos profile called up by the lan lan_profile for all IP data unless there is a
more relevant lan, profile, e.g. tftp.lan_profile (for tftp data).
21.1.3.2 Dial plan override of QOS profile
Specific QOS profiles can be selected for LAN packets associated with specific calls by specifying
the QOS profile to use in the dial plan dest statement, using the token QOS:. QOS: can be
specified for both calls being routed to the LAN and also for calls being received from the LAN.
The Vega does not use the same QOS values that it receives for an
incoming call in its responses for that call; the Vega must be
configured appropriately (manually) to use the correct QOS
settings.
For example, for a call being directed to the LAN:
dest =I F: 05, TEL: <1>, TA: 192. 168. 1. 4, QOS: 2
For a call being received from the LAN:
dest =I F: 02, TEL: <1>, QOS: 2
When overriding QOS profiles in the dial planner ensure that vl an_i d
is configured appropriately. Typically the vl an_i d should be the same
as the VoIP protocol specific vl an_i d because before a call is routed
(and hence before the QOS profile override takes over) there may be
ARPs or other messages between VoIP endpoints which must also be
routed through appropriately.
21.1.3.3 QOS profile configuration
21.1.3.3.1 Non 802.1 configuration
If the Vega is not configured for 802.1 operation then there are 4 configurable parameters in each
QOS profile:
[lan.if.x.8021q]
enabl e=0 ; di sabl e 802. 1 oper at i on
accept _non_t agged=1 ; accept non 802. 1 LAN packet s
; as wel l as 802. 1 packet s
[qos_profile.n]
name=def aul t
[qos_profile.n.tos]
def aul t _pr i or i t y=0 ; I P header TOS oct et
medi a_pr i or i t y=0 ; I P header TOS oct et
si gnal l i ng_pr i or i t y=0 ; I P header TOS oct et
The medi a_pr i or i t y is used for media packets, ie audio RTP packets and T.38 packets
NOTE
NOTE
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 337 - 6/2/2009
The si gnal l i ng_pr i or i t y is used for the VoIP signalling messages
The def aul t _pr i or i t y is used for any LAN traffic not associated with either call signalling or
call media (e.g. Telnet messages and Radius accounting messages).
21.1.3.3.2 802.1 configuration
If the Vega is configured for 802.1 operation then there are 9 configurable parameters in each
QOS profile:
[lan.if.x.8021q]
enabl e=1 ; enabl e 802. 1 oper at i on
accept _non_t agged=1 ; accept non 802. 1 LAN packet s
; as wel l as 802. 1 packet s
[qos_profile.n]
name=def aul t
[qos_profile.n.tos]
def aul t _pr i or i t y=0 ; I P header TOS oct et
medi a_pr i or i t y=0 ; I P header TOS oct et
si gnal l i ng_pr i or i t y=0 ; I P header TOS oct et
[qos_profile.n.8021q]
def aul t _pr i or i t y=0 ; 802. 1p pr i or i t y
medi a_pr i or i t y=0 ; 802. 1p pr i or i t y
si gnal l i ng_pr i or i t y=0 ; 802. 1p pr i or i t y
vl an_i d=0 ; 802. 1q Vi r t ual LAN I D
vl an_name=Def aul t
The medi a_pr i or i t y is used for media packets, ie audio RTP packets and T.38 packets
The si gnal l i ng_pr i or i t y is used for the VoIP signalling messages
The def aul t _pr i or i t y is used for any LAN traffic not associated with either call signalling or
call media (e.g. Telnet messages and Radius accounting messages).
The vl an_i d specifies the 802.1q Virtual LAN id to be used for LAN packets sent using this
profile. (All VoIP devices that need to communicate with each other must be configured with the
same VLAN id number.)
The vl an_name is provided for self-documentation purposes only. It does not affect the
information sent.
These items are configurable on the web browser interface on the QoS page select Modify
against the appropriate profile.
Vega Admin Guide R8.5 V1.5
Copyright VegaStream 2001-2009 - 338 - 6/2/2009
21.2 QoS Event Monitoring
Vegas may be configured to monitor certain QOS statistics, like jitter, buffer under / over flows
and packet loss. By monitoring their occurence against thresholds the Vega can provide alerts
when the thresholds are exceeded (and also when the problem recovers). Per-call and per-
gateway QOS events may be selected for monitoring.
For details on configuring QOS event monitoring in the Vega and details of the resulting
alarms, see information note IN 15 QOS Statistics
21.3 QoS Statistics Reports
Vegas can produce both per-call and per-gateway reports. These can be displayed either on
demand from an internal buffer, or delievered live to a terminal interface.
For details on configuring the Vega and the format of the resulting QOS statistics reports,
see information note IN 15 QOS Statistics
Vega Admin Guide R8.5 V1.5
APPENDIX A: SYSTEM EVENT LOG MESSAGES
System event log messages are created in the following format:
LOG: <t i me> <code ar ea gener at i ng msg>
( <ser i ousness>) R<r eason code>C<channel number > <message>
The following tables provide details of the reason codes and seriousness values. For further
details on reading LOG: messages, see section 9.
Reason Code
(and seriousness)
Reason Code
in Hex
Description
0-99 (Info)
0 00 Ent i t y/ ser vi ce st ar t i ng
1 01 I ncomi ng cal l
2 02 Out goi ng cal l
3 03 Connect cal l
4 04 Di sconnect
5 05 On- hook
6 06 Of f - hook
7 07 No r out e t o dest i nat i on
8 08 DSP l i cense l i mi t r eached
10 0A Fact or y def aul t s r est or ed
11 0B Rout e f ound
12 0C Ti me l oaded f r omser ver
15 0F Cal l bl ocked
16 10 Det ect ed syst emcl ock speed
17 11 conf i g par amet er wi t h ' aut o' set t i ng, has been set
t o def aul t , as appr opr i at e
18 12 conf i g. Par amet er wi t h i ncompat i bl e val ue has been
changed t o appr opr i at e set t i ng.
20 14 Pr of i l es r educed t o 40%of MAX when RAM < 16M
( V100 pr ot ot ypes)
21 15 Connect medi a
22 16 DHCP i t emdi scover y
23
17
17 Vega Reboot
24 18 Exceeded Max cal l s
25 19 Cal l congest i on on an i nt er f ace
17
watchdog and fatal reboots are reported in the log as <ser i ousness> Alert, user and coldstart
are <ser i ousness> Info
Vega Admin Guide R8.5 V1.5
Reason Code
(and seriousness)
Reason Code
in Hex
Description
100-150 (Warning)
100 64 No ser vi ces avai l abl e
101 65 No def aul t r out es
103 66 Cal l er I D r ecei ved af t er t he cal l has pr ogr essed
104 68 DSP channel r ef used
105 69 I SDN car d( s) f ai l ed
106 6A Ent i t y/ ser vi ce st oppi ng
108 6C DHCP di scover y f ai l ed
111 6F Bi l l i ng r ecor d l ost
112 70 Bi l l i ng l og appr oachi ng f ul l
113 71 Ent i t y message queue congest ed
114 72 TCP sessi on abor t ed ( keepal i ve t i meout )
115 73 Ent i t y message queue congest i on r el eased
116 74 Tone def i ni t i on not wr i t t en
117 75 I nval i d t one def i ni t i on
118 76 Too many t ones i n sequence
119 77 Tone i n sequence does not exi st
120 78 I nval i d t one sequence def i ni t i on
121 79 Tone sequence def i ni t i on not wr i t t en
122 7A I l l egal packet sour ce
123 7B SI P r egi st r at i on r econf i gur e i n unhandl ed st at e
124 7C DNS l ookup f ai l ed f or si p. def aul t _pr oxy
130 82 Mi smat ch of conf i gur ed l yr 1 set t i ngs ( Tel ogy 8
pr obl em)
140 8C Unabl e t o r ead conf i gur at i on
141 8D CALL_BLOCKED opt i on di sabl ed
142 8E I nval i d di al pl an conf i gur at i on - An endpoi nt can
onl y be assi gned t o one QoS pr of i l e
Reason Code
(and seriousness)
Reason Code
in Hex
Description
150-170 (Fail)
150 96 DSP boot code l oad f ai l ur e
151 97 DSP expect ed CODEC i mage absent
152 98 DSP boot code absent
153 99 DSP f ai l ur e
154 9A Open channel f ai l ur e det ect ed by r out er
155 9B SI P i ni t i al r esour ce al l ocat i on f ai l ur e
Vega Admin Guide R8.5 V1.5
Reason Code
(and seriousness)
Reason Code
in Hex
Description
si p. max_cal l s t oo l ar ge
156 9C Syst emFan Fai l ur e
157 9D I SDN car d f ai l ur e
170 AA Syst emOver heat / back t o nor mal t emper at ur e
Reason Code
(and seriousness)
Reason Code
in Hex
Description
171-190 (Alert)
171 AB Syst emi s r eady f or use
172 AC POTS i ncomi ng cal l
173 AD DSL act i ve
174 AE DSL i nact i ve
175 AF Cal l r ej ect ed; whi t el i st mat ch f ai l ed
176 B0 Cal l r ej ect ed; f i ndr out e f ai l ed
177 B1 Last act i ve cal l t er mi nat ed. New cal l s ar e bl ocked
178 B2 ' appl y' conf i gur at i on changes compl et e
179 B3 N channel s l i censed
180 B4 LAN act i ve
181 B5 LAN i nact i ve
182 B6 Gat ekeeper event
183 B7 An ' admi n' user has j ust l ogged i n
184 B8 Too many l ogi n f ai l ur es
185 B9 Passwor d changed f or user
186 BA Dupl i cat e MAC addr ess det ect ed
187 BB Boot - up scr i pt st at us r epor t i ng
188 BC Number of l i censed POTS por t s
189 BD Reboot due t o I P addr ess change by DHCP ser ver
190 BE VLAN val ues not pr eser ved
191 BF New cal l s unbl ocked
192 C0 QoS: Packet Loss bel ow t hr eshol d f or cal l number
193 C1 QoS: Packet pl ayout del ay bel ow t hr eshol d
194 C2 QoS: Packet j i t t er bel ow t hr eshol d
195 C3 QoS: Packet Loss t hr eshol d r eached
196 C4 QoS: Packet pl ayout del ay t hr eshol d r eached
197 C5 QoS: Packet j i t t er t hr eshol d r eached
198 C6 QoS: J i t t er buf f er over f l ow f or cal l r eached
199 C7 QoS: J i t t er buf f er under f l ow f or cal l
199 C7 QoS: I P Ser vi ce avai l abl e, LAN l i nk r est or ed
199 C7 QoS: I P Ser vi ce unavai l abl e due t o LAN f ai l ur e
199 C7 QoS: Packet pl ayout er r or r at e bel ow t hr eshol d f or
cal l
199 C7 QoS: Packet pl ayout er r or r at e t hr eshol d r eached
f or cal l
Vega Admin Guide R8.5 V1.5
Reason Code
(and seriousness)
Reason Code
in Hex
Description
199 C7 Syst emFan no l onger f ai l ed
Reason Code
(and seriousness)
Reason Code
in Hex
Description
200-255 (Error)
200 C8 No l ogi cal channel avai l abl e f or cal l
201 C9 H. 323 pr ef er r ed capabi l i t y not i n l i st
202 CA H. 323 f i r st capabi l i t y not G. 723. 1 or G. 729AnnexA
203 CB DSP i nt er nal er r or
204 CC Conf i gur at i on synt ax er r or
205 CD Dupl i cat e i nt er f ace i d f ound
206 CE Too many i nt er f aces r egi st er ed
207 CF Tone i ni t i al i sat i on f ai l ed
208 D0 Tone sequence i ni t i al i sat i on f ai l ed
209 D1 SI P WRI TE dat a t oo l ong
210 D2 I nval i d I SDN car d har dwar e ver si on f or T1 mode
211 D3 Compr essed web br owser page i s t oo bi g t o unpack
and di spl ay
255 FF Syst empower above t hr eshol d, r et ur ned bel ow
t hr eshol d.
Vega Admin Guide R8.5 V1.5
APPENDIX B: SIP SIGNALLING MESSAGES
The following SIP signalling messages are supported:
Vega FXS gateways can transmit INFO messages indicating a flash-hook event
Vega FXO gateways can receive INFO messages indicating a flash-hook event
Vegas can transmit and receive INFO messages indicating DTMF events
Vegas can receive INFO messages requesting playing of a tone (used to indicate call-
waiting)
Vegas can receive NOTIFY messages indicating if any voice messages are waiting
Vega FXS gateways can handle Alert-Info headers in an incoming INVITE (used for
generating distinctive ringing)
INFO Messages
INFO messages allow the Vega to:
1) Inform SIP clients that a flash hook event has occurred.
2) Inform SIP clients that a DTMF event has occurred.
3) Receive a request to play a DTMF tone.
4) Receive a request to play a tone (e.g.call-waiting).
The INFO messages contain a Content-Type field that will be in the form:
appl i cat i on/ si gnal l i ng_app_i d
where si gnal l i ng_app_id is defined by the si p. si gnal l i ng_app_i d configuration
parameter.
Vega Admin Guide R8.5 V1.5
INFO Messages DTMF and Hookflash MESSAGE
The generation of DTMF and Hookflash INFO messages requires the codec configured for out-
of-band DTMF and the Vega configured to send out INFO messages not just RFC2833.
check also parameters:
[_advanced.sip.info]
t x_hookf l ash
t x_dt mf
si p. dt mf _i nf o=mode1 (VegaStream standard):
Whenever a DTMF tone key is pressed on a POTS phone during a SIP call and the Vega
detects that tone, it will send a message like this:
I NFO si p: 3019775337@192. 168. 2. 175: 5060 SI P/ 2. 0
.
.
CSeq: 2 I NFO
Content-Type: application/signalling_app_id
Cont ent - Lengt h: xx
event DTMF 1 {key}
Where {key} is a single character indicating the key pressed ( 0, 1, 2 . . #, *)
When a hookflash event occurs, the Vega will send a message like this:
I NFO si p: 3019775337@192. 168. 2. 175: 5060 SI P/ 2. 0
.
.
CSeq: 2 I NFO
Content-Type: application/signalling_app_id
Cont ent - Lengt h: xx
event flashook
si p. dt mf _i nf o=mode2 (Cisco compatible):
Whenever a DTMF tone key is pressed on a POTS phone during a SIP call or a hookflash event
occurs, the Vega will send a message like this:
I NFO si p: 3019775337@192. 168. 2. 175: 5060 SI P/ 2. 0
.
.
CSeq: 2 I NFO
Cont ent - Type: appl i cat i on/ dt mf - r el ay
Cont ent - Lengt h: xx
Signal {key}
Duration 250
Where {key} is a single character indicating the key pressed ( 0, 1, 2 . . #, *) , a hookflash
is indicated by {key} being the ! character.
Duration is always given as 250ms.
Vega Admin Guide R8.5 V1.5
INFO Messages PLAY TONE MESSAGES
When the remote end wants the Vega to play a tone, it can activate this by sending a message
like this:
I NFO si p: 3019775337@192. 168. 2. 175: 5060 SI P/ 2. 0
.
.
CSeq: 2 I NFO
Content-Type: application/signalling_app_id
Cont ent - Lengt h: xx
play tone preset 1
INFO message body Configuration
pl ay t one pr eset 1
Or :
pl ay t one Cal l Wai t i ngTone1
t one def i ned by t ones. cal l wai t 1_seq
pl ay t one pr eset 2
Or :
pl ay t one Cal l Wai t i ngTone2
t one def i ned by t ones. cal l wai t 2_seq
E.g. for call waiting tone 1:
admi n >show t ones. cal l wai t 1_seq
[tones]
cal l wai t 1_seq=6
This points to the definition of tone sequence 6:
admi n >show t ones. seq. 6
[tones.seq.6]
name=cal l wai t 1_seq
r epeat =0
[tones.seq.6.tone.1]
dur at i on=350
pl ay_t one=7
NOTIFY Messages
NOTIFY messages allow the Vega to receive notification of waiting voice messages.
NOTI FY si p: 3019775337@192. 168. 2. 153 SI P/ 2. 0
.
.
Cseq: 1 NOTI FY
Content-Type: text/plain
Cont ent - Lengt h: xx
Messages-Waiting: mw
Vega Admin Guide R8.5 V1.5
Where mw can be:
yes
no
n where n=0, 1, 2, . . . and specifies the number of waiting messages
When the Vega receives a message where n>0 or mw is yes, then the Vega will:
1) Play a "stutter" dial-tone to the POTS user next time he/she takes the phone off-hook.
2) Send an MWI (message waiting indication) signal to the phone.
NOTE
1. The stutter dial-tone is specified by
t ones. st ut t er d_seq. This defines which tone
sequence to use as the stutter dial-tone.
By default:
[tones]
st ut t er d_seq=2
2. To send an MWI signal to the phone, the Vega uses
FSK tones. Some phones require a short voltage
drop before the sending of the tones (like a
hookflash) this is not supported.
INVITE Messages with Alert-Info
Vega FXS gateways can handle INVITE messages containing an "Alert-Info" field. The Alert-
Info header will look something like this:
Al er t - I nf o: bel l cor e- r 1
The Vega will try to match up the alert type (in this case, "bel l cor e- r 1") to an
_advanced. pot s. r i ng. x. name field in the configuration.
In this case, there would be a match with the following entry:
[ _advanced. pot s. r i ng. 4]
name=bel l cor e- r 1
f r equency=20
r epeat =1
r i ng1_on=350
r i ng1_of f =350
r i ng2_on=900
r i ng2_of f =300
r i ng3_on=350
r i ng3_of f =3700
LIMITATIONS: This currently only works on calls on POTS interfaces that are in group 1, e.g.
pot s. por t . 3. i f . 1
When NO "Alert-Info" field is present, then the Vega FXS port will use the ring specified by:
Vega Admin Guide R8.5 V1.5
pot s. por t . x. i f . 1. r i ng_i ndex where x (1-8) is the called POTS
interface.
If the "Alert-Info" field is present, then the Vega will try to use the ring specified.
INVITE Message Session Description
Some systems require the c= line to be in in the SDP media description, others require it in the
SDP session description. Vegas can support either requirement based on the configuration of
the parameter:
_advanced. si p. sdp. sess_desc_connect i on=0
the c= line appears in the SDP media description. For example:
v=0
o=Vega50 7 1 I N I P4 136. 170. 208. 245
s=Si p Cal l
t =0 0
m=audi o 10012 RTP/ AVP 0
c=I N I P4 136. 170. 208. 245
a=r t pmap: 0 PCMU/ 8000
_advanced. si p. sdp. sess_desc_connect i on=1
the c= line appears in the SDP session description. For example:
v=0
o=Vega50 8 1 I N I P4 136. 170. 208. 245
s=Si p Cal l
c=I N I P4 136. 170. 208. 245
t =0 0
m=audi o 10014 RTP/ AVP 0
a=r t pmap: 0 PCMU/ 8000
Vega Admin Guide R8.5 V1.5
APPENDIX C: DTMF TONE FREQUENCIES
Frequency (Hz) 1209Hz 1336Hz 1477Hz 1633Hz
Frequency (Hex) 0x4b9 0x538 0x5c5 0x661
697Hz 0x2b9 1 2 3 A
770Hz 0x302 4 5 6 B
852Hz 0x354 7 8 9 C
941Hz 0x3ad * 0 # D
Vega Admin Guide R8.5 V1.5
APPENDIX D: HEXADECIMAL TO DECIMAL CONVERSION
Hex Dec Hex Dec Hex Dec Hex Dec Hex Dec Hex Dec Hex Dec Hex Dec
00 0 20 32 40 64 60 96 80 128 A0 160 C0 192 E0 224
01 1 21 33 41 65 61 97 81 129 A1 161 C1 193 E1 225
02 2 22 34 42 66 62 98 82 130 A2 162 C2 194 E2 226
03 3 23 35 43 67 63 99 83 131 A3 163 C3 195 E3 227
04 4 24 36 44 68 64 100 84 132 A4 164 C4 196 E4 228
05 5 25 37 45 69 65 101 85 133 A5 165 C5 197 E5 229
06 6 26 38 46 70 66 102 86 134 A6 166 C6 198 E6 230
07 7 27 39 47 71 67 103 87 135 A7 167 C7 199 E7 231
08 8 28 40 48 72 68 104 88 136 A8 168 C8 200 E8 232
09 9 29 41 49 73 69 105 89 137 A9 169 C9 201 E9 233
0A 10 2A 42 4A 74 6A 106 8A 138 AA 170 CA 202 EA 234
0B 11 2B 43 4B 75 6B 107 8B 139 AB 171 CB 203 EB 235
0C 12 2C 44 4C 76 6C 108 8C 140 AC 172 CC 204 EC 236
0D 13 2D 45 4D 77 6D 109 8D 141 AD 173 CD 205 ED 237
0E 14 2E 46 4E 78 6E 110 8E 142 AE 174 CE 206 EE 238
0F 15 2F 47 4F 79 6F 111 8F 143 AF 175 CF 207 EF 239
10 16 30 48 50 80 70 112 90 144 B0 176 D0 208 F0 240
11 17 31 49 51 81 71 113 91 145 B1 177 D1 209 F1 241
12 18 32 50 52 82 72 114 92 146 B2 178 D2 210 F2 242
13 19 33 51 53 83 73 115 93 147 B3 179 D3 211 F3 243
14 20 34 52 54 84 74 116 94 148 B4 180 D4 212 F4 244
15 21 35 53 55 85 75 117 95 149 B5 181 D5 213 F5 245
16 22 36 54 56 86 76 118 96 150 B6 182 D6 214 F6 246
17 23 37 55 57 87 77 119 97 151 B7 183 D7 215 F7 247
18 24 38 56 58 88 78 120 98 152 B8 184 D8 216 F8 248
19 25 39 57 59 89 79 121 99 153 B9 185 D9 217 F9 249
1A 26 3A 58 5A 90 7A 122 9A 154 BA 186 DA 218 FA 250
1B 27 3B 59 5B 91 7B 123 9B 155 BB 187 DB 219 FB 251
1C 28 3C 60 5C 92 7C 124 9C 156 BC 188 DC 220 FC 252
1D 29 3D 61 5D 93 7D 125 9D 157 BD 189 DD 221 FD 253
1E 30 3E 62 5E 94 7E 126 9E 158 BE 190 DE 222 FE 254
1F 31 3F 63 5F 95 7F 127 9F 159 BF 191 DF 223 FF 255