211 MPLS Te

SOCIETY FOR SCIENCE AND EDUCATION
UNITED KINGDOM
TRANSACTIONS ON
TNC NETWORKS AND COMMUNICATIONS
V OLUME 3, I SSUE 5
ISSN: 2054 -7420
Links Utilization in MPLS Networks Operating with Traffic

Engineering Signal Protocols
Ghufran Saady Abd-almuhsen1, Mahmoud M. Al-Quzwini2, and Raad Sami. Fyath3
Department of Computer Engineering, Al-Nahrain University, Baghdad, Iraq
1
ghufransaady@yahoo.com, 2quzwini72@yahoo.com, 3rsfyath@yahoo.com
ABSTRACT
Multiprotocol Label Switching (MPLS) is a technology that ensures efficient transmission with high speed
and lower delays. Traffic Engineering (TE) signal protocols are usually used for active management of the
MPLS networks for efficient utilization of resources. This paper presents performance investigation of
MPLS TE signal protocols to get a guideline to utilize transmission links efficiently. Comparison is made
between two TE signal protocols, namely Resource Reservation Protocol (RSVP) and Constraint-based
Routing Label Distribution Protocol (CR-LDP). Simulation results are presented for three MPLS networks
having different topologies, which are implemented in OPNET (version 14.5) environment to support
different applications in the absence and presence of quality of service (QoS) algorithms. The results
reveal that MPLS network with CR-LDP TE signal protocol has better performance in the term of link
utilization. The RSVP reserves certain paths for transmission while the CR-LDP utilizes almost all of the
available links.
Key words: MPLS Networks, Traffic Engineering Signal Protocols, Quality of Service.
1 Introduction
Multi Protocol Label Switching is raised from the Internet Engineering Task Force’s (IETF) effort to
standardize a number of proprietary multilayer switching solutions that were initially proposed in the mid-
1990s. MPLS integrates layer 3 (routing) and layer 2 (switching) functionalities [1]. MPLS introduces
connection-oriented forwarding model by replacing the routing of IP packets based on the IP header
information with the short four-byte label-based switching, as shown in Figure (1).
The mechanism does not build forwarding decision based on the traditional destination IP address on
sophisticated lookup routing table. This fixed-length switching concept is to some extent similar to that
used in ATM and Frame Relay networks, and it is independent of the used layer 2 technologies. MPLS has
been designed to provide an admirable solution to present shortcomings of IP routing in the area of Traffic
Engineering (TE), Quality of Service (QoS), Virtual Private Networks (VPN) and Differentiated Services
(DiffServ) [2]. In comparison of DiffServ with MPLS which is evolving as a futurity protocol. MPLS is
desirable over DiffServ since it utilizes “Multi Protocol Architecture” depending on simple label switching
technique. Traffic can be simply differentiated thereby ensuring QoS based on traffic types. Applications
like VPNs need MPLS to achieve high quality end-to-end service. The new and better network topologies-
DOI: 10.14738/tnc.35.1448
Publication Date: 14th September 2015
URL: http://dx.doi.org/10.14738/tnc.35.1448
Ghufran Saady Abd-almuhsen, Mahmoud M. Al-Quzwini and Raad Sami. Fyath; Links Utilization in MPLS Networks
Operating with Traffic Engineering Signal Protocols, Transactions on Networks and Communications, Volume 3
No. 5, October (2015); pp: 1-27
Any Transport over MPLS (AToM), MPLS over Voice over Internet Protocol (VoIP) and video traffics etc.
have resulted in perceivable QoS [3].
Figure (1): MPLS forwarding label between routers. LSR denotes Label Switching Router
MPLS enabled network can provide efficient TE services, flow based services, better traffic shaping, etc.
In [4], routing based traffic flow shaping is introduced, where total traffic is split over multiple Label
Switching Paths (LSPs) in the MPLS network. This method is powerful for solving some of the routing
problems like mismatch and bottleneck problems. Network with MPLS can propose good QoS to delay
critical traffic such as meetings, VoIP and video conference. Network failure such as crash of network
elements, link faults or congestion are easily managed in MPLS networks [5].
MPLS can be considered a technology to forward the packets in IP intangible networks. The Entire MPLS
network can be split into two parts namely MPLS edge and MPLS core [6]. MPLS edge is the border of the
MPLS network consisting of egress and ingress routers. MPLS core bound intermediate Label Switching
Routers (LSRs), through which Label Switched Paths (LSPs) are established [7].
A traffic engineering matter in the Internet consists of setting up paths between the edge routers in a
network to meet traffic needs while attaining low congestion and improving the utilization of network
resources. Practically, the usual key purpose of traffic engineering is to eliminate the utilization of the
most heavily used links in the network, or the maximum of link utilization. As the maximum link utilization
qualitatively reveals that congestion sets in when link utilization rises higher, and hence it is necessary to
eliminate the link utilization throughout the network such that no bottleneck link occurs. It is known that
this problem of reducing the maximum link utilization can be achieved by the multi-commodity network
flow formula of powerful routing, which leads to dividing traffic over multiple paths between source and
destination pairs [8]. This paper addresses link utilization in MPLS-based networks incorporating TE signal
protocols and Qos algorithms.
2 Related Work
In 2010, Shyry and Ramachandran [9] discussed the effect of MPLS on network performance with the aid
of the Nash equilibrium algorithm. The results show that optimized performance can be obtained by
2
Transactions on Networks and Communications; Volume 3, Issue 5, October 2015
reducing the latency and raising the link utilization. But rising link utilization leads to a gradually increase
of the congestion in the network. To overcome this specific problem, a formulation called dual
programming formulation was performed which has group of constraints that have to be satisfied along
with Open Shortest Path First (OSPF) protocol so as to eliminate the maximum link utilization.
In 2011, Pelsser and Bonaventure [10] discussed the Service Provider's (SP's) requirements for the
utilization of MPLS LSPs across Autonomous System (AS) boundaries. A minimum set of extensions was
introduced to Resource Reservation Protocol-Traffic Engineering (RSVP-TE) that allows setting inter-AS
LSPs in accordance with the Service Provider requirements. The results show how LSP protection
techniques can be extended to provide links or node failure protection for the border routers and inter-
AS.
In 2012, Bongale [5] compared link utilization among networks running Routing Information Protocol
(RIP), Open Shortest Path First (OSPF) and MPLS. The results showed that networks configured with OSPF
and RIP routing technicalities are not capable of managing the incoming traffic efficiently. When the
network traffic increases, shortest path from source node to destination node is heavily congested and
lead to loss of transmitted data, while MPLS is capable of handling incoming traffic effectively by
portioning the traffic over unutilized links. This will ensure packets, that entering into MPLS core, to reach
the destination with minimum queuing delay. The results also indicate that MPLS-TE is most appropriate
for enormous traffic volume. OPNET simulator was used to get the results and performance was compared
considering data consisting of voice traffic and web browsing only.
In 2012, Aziz et al. [11] presented a QoS performance study of real-time applications such as video
conferencing and voice in terms of Packet Delay Variation (PDV) over DiffServ in the absence and presence
of MPLS-TE in IPv4/IPv6 networks using OPNET simulator. The interaction of Assured Forwarding (AF)
traffic aggregation, Expedited Forwarding (EF), link congestion, in addition to the effect of performance
metric like PDV were also studied. The performance of DiffServ and MPLS-TE combination in IPv4/IPv6
network was elucidated and analyzed. The results show that IPv6 encounter more PDV than their IPv4
counterparts.
In 2013, Bhandure et al. [12] studied MPLS and Non-MPLS networks and presented an overview of the
MPLS technology and related IETF standards. The results show that MPLS is faster and has better
performance than traditional IP routing. Performance was compared by observing parameters such as
number of transmitting received packets, Jitter (delay variation) and end-to-end delay. GNS 3.0 simulator
was used to simulate the networks. The simulations were setup using a traditional IP network without TE
(composed of OSPF and BGP) and MPLS network (composed of OSPF and BGP).
In 2013, Ibrahim [13] discussed the performance of MPLS-TE signal protocols, namely the Resource
Reservation Protocol (RSVP) and Constraint based Routed–Label Distribution Protocol (CR-LDP), with
different applications including voice, video and data. Performs evaluation of the two protocols shows
that CR-LDP outperforms the RSVP in terms of response time and the average transmitted and received
packets in all applications. The link utilization capability of these protocols was also addressed with
different transmission loads.
In 2014, Sulaiman and Alhafidh [14] discussed the performance analysis of multimedia traffic over MPLS
communication networks with TE. The performance metric ofMPLS-TE and IP model networks was
compared. The compared parameters were end-to-end delay, delayvariation, packet send and receive,
Copyright © Society for Science and Education United Kingdom 3

No. 5, October (2015); pp: 1-27
File Transfer Protocol (FTP) responsetime. The results show that MPLS-TE performance is better than
traditional IP network model.
Most of the previous works focuse on traffic performance comparison between both MPLS and Non-MPLS
networks by using simulation tools. This paper focuses mainly on the performance of MPLS networks
when TE is taking into account with some signaling protocols (CR-LDP and RSVP) and QoS algorithms.
Emphasize is being placed on the key role played by link utilization.
3 Background
This section introduces brief description of MPLS network and TE signal protocols.
3.1 MPLS Network

In MPLS, packets are sent to their destinations by labeling them and forwarding them. Short, fixed-length
labels are added to the IP packets when they enter the network. Consequently, instead of using the IP
header information, the labels are used to forward the packets to their destinations. To do this, a new
protocol is developed for classifying the labels. Extensions to existing protocols are also used to ease this
[15].
MPLS can be considered a technology to forward the packets in IP intangible networks. The Entire MPLS
network can be split into two parts namely MPLS edge and MPLS core [16]. MPLS edge is the border of
the MPLS network consisting of egress and ingress routers shown in Figure (2). MPLS core bound
intermediate Label Switching Routers (LSRs), through which Label Switched Paths (LSPs) are established
[16].
Figure (2): MPLS domain network
General terms correlating with MPLS network and their explanation are specified in the following
• Label Description: A short, fixed length packet identifier.
• Label Edge Router (LER): A device that operates at the edge of the access network and MPLS
network.
• Label Switching Router : A router which is located in the MPLS domain and forwards the packets
based on label switching.
• Forward Equivalence Class (FEC) : A description of a group of packets sharing the same transport
requirements.
• Label Switched Path (LSP) : A route established between two Label Edge Router (LER) which work
as a path for forwarding labeled packets over LSPs.
4
The two planes , namely control plane and data plane, highlight the operation of MPLS are shown in
Figure (3) [17].
Figure (3): MPLS planes [17]
3.2 Traffic Engineering Signal Protocols

With the standardization of MPLS by IETF, traffic engineering obtained its popularity due to the supported
features of the MPLS for Traffic Engineering more than the traditional IP networks. The main structure
blocks of the MPLS Traffic Engineering Model are Path Management, Network State Information
Dissemination, Traffic Assignment and Network Management [16]. Due to the online and offline usability
of the MPLS network and the capability to list at any point of time, Traffic Engineering has gained its
popularity [18].The packets in MPLS network are forwarded using the level swapping. This forwarding of
packets gives more control for expeditiously forwarding packets [17]. Figure (4) shows the relation
between Interior Gateway Protocol (IGP), MPLS and constraint-based routing for the path selection in a
network. The path selection procedure depends on the availability of the protocols. In absence of MPLS,
the path selection is done by IGP and in the presence of the MPLS the path selection or the signaling
protocols of the MPLS are implemented [19].
Figure (4): Interaction of the various components of an MPLS-based Traffic Engineering solution [19].
Traditionally IP packets were forwarded looking into its destination address at every router in the path.
The packets were forwarded based on the shortest path metric, which is the cost calculated using the
time it takes to reach the next hop. When the traffic in the network increases, the link with shortest path
becomes heavily congested while the links with higher paths are underutilized resulting in the uneven
loads in the links available, on the cost of traffic resources. The development of MPLS addresses these

No. 5, October (2015); pp: 1-27
problems with the use of constraint based routing (CBR). The dynamic use of the measuring tools and the
accountability of all the possible multiple paths and its characteristics (bandwidth, policy and topology)
by CBR makes it easier for implementation of Traffic Engineering efficiently [17].
Signaling protocols are used to set up the paths for the packets to follow, these paths are usually known
as Label Switched Path. There are many protocols which can be used for paths selection, but here only
the signaling protocols that support Traffic Engineering are explained [18]. In this paper, two types of
signal protocols are used, Constraint-Based Label Distribution Protocol (CR-LDP) and Resource
Reservation Protocol (RSVP).
3.3 Constraint-Based Label Distribution Protocol

Label Distribution Protocol (LDP) is designed by a working group at IETF from the ground up for the
particular reason for distributing MPLS labels, consequently setting up LSPs in the MPLS domain. LDP
works closely with IGP routing protocols and is usually called "hop-by-hop" forwarding. LDP does not
support TE because it always chooses the same physical path that traditional IP routing would select. The
reason behind setting up an LSP that follows the same path as traditional IP instead of just using traditional
IP routing was primarily to accelerate the forwarding in routers. In traditional IP routing the next hop for
each packet is found by a longest match prefix lookup on the IP header in the routing table. These lookup
could in some cases, where the routing tables were large, be time consuming and it was surmised that
data forwarding with label switching instead of IP lookups would speed up data forwarding. However, the
forward speed of IP packets is not a matter anymore. Because of the development in routing technology,
LDP is not frequently used for label distribution nowadays. There is an extension to the original LDP
protocol that presents the new functionality of the LDP protocol called CR-LDP [20].
CR-LDP is an extent of LDP to support constraint based routed LSPs. The term constraint implies that in a
network and for each group of node there exists a group of constraint that must be satisfied for the link
or links between two nodes to be selected for an LSP. LSRs that use CR-LDP to interchange label and FEC
mapping information are called LDP peers, they interchange this information by forming a LDP session
[21].
3.4 Resource Reservation Protocol

An alternate signaling protocol to LDP and CR-LDP is the resource reservation protocol traffic engineering
(RSVP-TE). RSVP-TE is an extension of the resource reservation protocol which was designed to support
the integrated services (intserv) architecture. The intserv architecture was improved by IETF in the mid
1990s with a view to introducing QoS in the IP network.
The following two service classes were defined in intserv [17], [22]
i. Guaranteed service: This service provides firm bounds on the end-to-end queuing delay with no
packet loss for all conforming packets.
ii. Controlled-load service: This service provides the user with a QoS that closely approximates the
QoS of the best effort service that the user would receive from an unloaded network. Specifically,
a user might assume the following:
6
• A very high percentage of transmitting packets will be successfully transported by the network to
the receiver. The percentage of packets not successfully transported must widely approximate
the basic packet error rate of the transmission links
• The end-to-end delay experienced by a very high percentage of the transported packets will not
greatly achieve the minimum end-to-end delay experienced by any successfully transported
packet.
RSVP is soft state protocol, which means that when a path has been setup by RSVP it has to be regularly
updated to keep the recourses reserved. The requests for reservation are made from the receiver end of
the path, so RSVP is a receiver-oriented protocol. When RSVP is used for LSP setup the ingress router
starts by sending a PATH message on the path where an LSP will be set up. Each transportation router on
that path has to exam if it has the possibility to set up the requested LSP. If the requested LSP is discard,
an error message is returned upstream until it reaches the ingress router. Furthermore the path message
is sent to the next transportation router in the path until it arrives the egress router [23].
4 Simulated Network Topologies and Setup Parameters

This section presents a full description of network setup and topologies that used in the simulation. An
investigation of MPLS Traffic Engineering signal protocols capabilities to utilize the transmission link in the
absence and presence of QoS is also presented. Different applications including voice, video, File Transfer
Protocol (FTP), Telnet, print, Electronic mail (E-mail), database and Hyper Text Transfer Protocol (HTTP)
are used for the performance evaluation. The parameters that are considered throughout the study are
Throughput (the average number of bits successfully transmitted or received by the transmitter or the
receiver channel per unit time, in bits per sec) and Link utilization (the percentage of the throughput to
the data rate of the link used in the transmission).
The simulation environment used in this work is based on Optimum Network Engineering Tool (OPNET)
14.5 simulator. OPNET is a real-time simulator suitable mainly for the design and analysis of network
models. The VoIP traffic is sent between the workstation (voice 1) and work station (voice 2). The same
terminology is followed with video traffic, which is sent between the workstation (video 1) and
workstation (video 2). For other applications, (HTTP, FTP, Database, E-mail, Print and Telnet), the traffic is
sent between workstations and servers.
4.1 Network Topologies

The following parameters are considered in the simulation
(i) DS1 links (data rate 1.544 Mbps) are used in the core network.
(ii) Network applications are divided into low, medium and high loads.
Figures (5a)-(5c) show the topologies of the three simulated MPLS networks using the two TE signal
protocols ( RSVP and CR-LDP). The first network has six routers, two LER routers and four LSR routers as
shown in Figure (5a). In the second network, the number of routers increases. The network uses
eight LSR routers and two LER routers, as shown in Figure (5b). The third network has same number of
routers as the second network but with increasing number of links (38 links) as shown in Figure (5c).

No. 5, October (2015); pp: 1-27
Table (1): Numbers of elements used in the simulated three networks.
Number of Elements
Networks
LSR Router LER Router Link
First 4 2 20
Second 8 2 28
Third 8 2 38
(a)
Figure (5): MPLS network topologies used in the simulation.
(a) First network (b) Second network (c) Third network.
(b)
8
(c)
Figure (5): continued.
4.2 Parameters used in the Simulation

Table (2) shows the parameters of the applications to be used during the simulations. Voice work stations
use G.711 codec with data rate of 64 kbps. There are many factors that indicate the quality of voice which
include choice of codec, packet loss, and delay. For VoIP applications, it is required that end-to-end packet
delay shouldn’t exceed 150ms in order that the quality of the established VoIP call is acceptable [25].
The voice delay in G.711 can be divided into two contributing components, which are explained as follows
(i) The delay provided by the G.711 codec for encoding and packetization is 1 and 24ms,
respectively. Therefore, the delay at the transmitter according to these two delays with
compression is approximated to a fixed delay of 25 ms.
(ii) At the receiver, the delay comes from decompression, buffering, playback, and
depacketization delay. The total delay due to these factors mentioned above is approximated
to a fixed delay of 45 ms.
Table (2): Applications parameters

VoIP Applications
Encoder Scheme G.711 (PCM)
Type of Service Interactive Voice
Video Applications
Frame rate 10 frames/sec
Frame Size 128*120 pixels
Type of Service Streaming Multimedia
FTP Applications
File Size 5000 byte
Type of Service best effort
HTTP Applications
HTTP Specification HTTP 1.1
Page Properties 1000 byte
Print Applications (color file)
File size 3000-90000 byte
Type of Service Best effort

No. 5, October (2015); pp: 1-27
Telnet Applications
Terminal Traffic (normally Mean =144, variance=60
distributed) byte
E-Mail Applications
E-Mail Size 2000 byte
Database Applications
Transaction Size 512 byte
The maximum acceptable network delay can be determined from the above transmitter and receiver
delays to be 80 ms nearly (150-25-45) ms, where the 150 ms represents the maximum acceptable end-to-
end delay, so that the quality of the established VoIP call is acceptable [26]. Video workstations transfer
10 frames per second (sec); each frame be formed of 128x120 pixels. FTP work stations used files of size
5000 bytes [27]. HTTP work stations use pages of size 1000 bytes, HTTP 1.1 is employed in this study, HTTP
1.1 is a revision of the original HTTP (HTTP 1.0). In HTTP 1.0 a separate connection to the same server is
made for every resource request. HTTP 1.1 can reuse a connection multiple times to download images,
page inter-arrival times are exponentially distributed with mean 60 sec [28]. Emails are sent with inter-
arrival times exponentially distributed with mean 360 sec. For database application, the transactions
arrive with inter-arrival times exponentially distributed with mean 12 sec. Telnet application initiates
commands from terminal to telnet host, these commands consist of a normally distributed amount of
bytes with mean 144 and variance 60. Best-effort service means that the user obtains unspecified variable
bit rate and delivery time, depending on the current traffic load of the network [26].
5 Link Utilization in the Absence of QoS

In this section, the link utilization of the three MPLS networks operating with CR-LDP and RSVP under
different load conditions (low, medium and high) are studied and compared. No Qos algorithms are
applied here. The term uplink refers to the paths carrying data from the work stations side to the server’s
side and the term downlink refers to the paths carrying data from the server’s side to the work stations
side.
The first MPLS Network used in the simulation is shown in Figure (5a) and examined here under low-load
condition. Figures (6a) and (6b) show the link utilization for the uplink paths and down link paths of the
MPLS network with CR-LDP TE signal protocol, respectively. Figures (6c) and (6d) present the uplink and
downlink utilization of the MPLS network with RSVP TE signal protocol, respectively.
Tables (3)-(5) list the average link utilization of all paths when the two TE signal protocols are used in the
three networks, respectively. Results are given for the three load conditions (low, medium, and high). The
numbers given in the tables are the time averages of the results obtained in the simulation (such that
given in Figure (6)).
10
Figure (6a): Link utilization of the uplink paths of network 1 with CR-LDP TE signal protocol (low load)
Figure (6b): Link utilization of the downlink paths of Network 1 with CR-LDP TE signal protocol (low load).
Figure (6c): Link utilization of the uplink paths of Network 1 with RSVP TE signal protocol (low load).

No. 5, October (2015); pp: 1-27
Figure (6d): Link utilization of the downlink paths of Network 1 with RSVP TE signal protocol (low load).
Table (3): Average link utilization of Network 1 operating with CR-LDP and RSVP TE signal protocols in the
absence of QoS.
Utilization %
Link Low load Medium load High load
CR-LDP
RSVP CR-LDP RSVP CR-LDP RSVP
14→20 3.20 92.03 18.80 11.42 18.80 11.42
20 →14 89.35 2.05 35.32 59.68 35.32 59.68
18→15 46.06 0.00 92.03 8 92.03 8
15→18 1.07 92.04 3.90 93.37 3.90 93.37
18→21 0.43 90.06 2.39 33.26 2.39 33.26
21→18 45.55 0.00 91.86 6.49 91.86 6.49
19→15 49.29 92.07 4.28 87.19 4.28 87.19
15→19 93.34 0.00 92.02 1.86 92.02 1.86
19→18 0.00 0.00 0 0 0 0
18→19 0.00 92.22 0 0 0 0
20 →18 2.18 0.00 14.90 0 14.90 0
18→20 0.00 1.99 1.52 59.69 1.52 59.69
20→19 1.10 92.02 3.91 11.01 3.91 11.01
19→20 89.34 0.00 34.80 93.32 34.80 93.32
21→14 4.46 90.17 59.61 33.79 59.61 33.79
14→21 93.34 0.00 59.61 1.36 59.61 1.36
21→19 48.30 0.00 0.40 87.01 0.40 87.01
19→21 4.06 92.04 57.23 0 57.23 0
21→20 0.00 0.00 0 0 0 0
20→21 0.00 0.00 0 0 0 0
12
Table (4): Average link utilization of Network 2 in the absence of QoS.

Utilization %
Link Low load Medium load High load
CR-LDP RSVP CR-LDP RSVP CR-LDP RSVP
14→23 37.58 11.68 37.58 11.68 0.62 92.64
23→14 6.58 71.86 6.58 71.86 88.66 0.00
20→19 0.00 0.00 0.00 0.00 0.00 0.00
19→20 0.00 0.00 0.00 0.00 0.00 0.00
20→25 0.43 0.00 0.43 0.00 0.00 0.00
25→20 6.58 17.95 6.58 17.95 3.46 0.00
20→26 56.43 83.51 56.43 83.51 80.87 0.00
26 →20 12.51 0.00 12.51 0.00 0.33 92.62
21→19 0.00 0.00 0.00 0.00 0.00 0.00
19→21 0.00 0.00 0.00 0.00 0.00 0.20
21→20 56.85 83.52 56.85 83.52 80.87 0.18
20→21 87.30 18.07 87.30 18.07 3.77 92.62
21→24 0.00 0.00 0.00 0.00 0.00 0.00
24→21 0.00 0.00 0.00 0.00 0.00 0.00
22→14 87.30 19.88 87.30 19.88 3.86 92.64
14→22 58.02 84.82 58.02 84.82 92.05 16.04
22→21 56.86 83.52 56.86 83.52 80.92 0.18
21→22 58.02 18.10 58.02 18.10 3.86 92.62
22→24 1.17 1.31 1.17 1.31 11.14 15.86
24→22 0.00 1.80 0.00 1.80 0.00 0.24
23→22 0.00 0.00 0.00 0.00 0.00 0.00
22→23 0.00 0.00 0.00 0.00 0.00 0.00
23→24 37.58 11.69 37.58 11.69 0.62 92.64
24→23 6.58 72.29 6.58 72.29 89.08 0.00
24→19 38.74 12.98 38.74 12.98 11.75 92.77
19→24 6.58 74.07 6.58 74.07 89.08 0.24
25→15 39.17 12.98 39.17 12.98 11.79 92.77
15→25 92.43 92.04 92.43 92.04 92.62 0.44
25→19 6.59 74.1 6.59 74.1 89.19 0.00
19→25 38.74 12.98 38.74 12.98 11.88 92.77
26→15 56.42 83.09 56.42 83.09 80.44 0.00
15→26 12.51 0.00 12.51 0.00 0.00 92.63
26→25 0.00 0.00 0.00 0.00 0.00 0.00
25→26 0.00 0.00 0.00 0.00 0.00 0.00

No. 5, October (2015); pp: 1-27
Table (5): Average link utilization of Network 3 in the absence of QoS.
Utilization %
LINK Low load Medium load High load
RSVP CR-LDP RSVP CR-LDP RSVP CR-LDP
12→18 0.00 6.95 14.28 8.16 21.80 92.29
18→12 93.07 92.18 5.68 43.68 0.00 4.16
12→20 93.07 0.00 93.07 0.00 0.00 13.34
20→12 0.00 7.15 2.30 1.00 2.20 2.82
12→21 0.00 0.00 0.00 1.00 93.06 1.18
21→12 2.22 0.88 1.03 3.41 1.74 91.86
12→24 0.00 92.50 5.68 43.95 0.00 4.14
24→12 0.00 6.95 14.28 8.17 21.80 91.98
14→17 93.08 0.00 0.17 59.60 93.37 22.22
17→14 0.00 0.88 0.00 0.00 0.00 6.74
14→22 1.20 0.50 93.09 1.09 4.44 5.52
22→14 12.26 63.90 93.06 92.18 0.00 25.22
14→23 0.00 12.87 12.37 44.05 0.00 4.16
23→14 93.06 29.59 14.28 8.28 21.80 92.29
14→26 93.07 92.50 5.68 0.00 0.00 0.00
26→14 0.00 7.03 0.00 0.00 0.00 0.00
17→18 0.00 0.00 93.09 60.14 93.37 25.25
18→17 0.00 0.00 0.00 92.58 0.00 18.40
18→20 1.20 6.13 0.00 0.00 0.00 0.00
20→18 12.26 92.51 0.00 0.00 0.00 0.00
21→20 0.00 0.00 0.00 0.00 0.00 6.79
20→21 0.00 0.00 93.09 60.16 93.36 22.58
21→23 12.26 64.11 0.00 1.00 0.00 1.19
23→21 1.20 0.50 1.03 3.37 1.74 92.29
22→17 0.00 0.00 0.00 0.00 1.74 0.00
17→22 2.22 1.00 0.00 0.00 0.00 0.00
22→24 0.00 0.00 1.02 3.43 0.00 92.29
24→22 0.00 0.00 0.00 1.00 93.06 1.20
23→17 2.22 1.00 0.00 0.00 0.00 0.00
17→23 0.00 0.00 0.00 0.00 0.00 0.00
23→26 0.00 0.00 0.00 0.00 0.00 0.00
26→23 0.00 0.00 0.00 0.00 0.00 0.00
24→18 0.00 0.00 0.00 0.00 2.20 0.00
18→24 0.00 0.00 0.00 0.00 0.00 0.00
26→20 0.00 7.38 2.30 1.00 0.00 2.84
20→26 93.07 0.00 93.07 0.00 0.00 13.62
26→21 0.00 5.78 10.08 0.00 0.00 2.83
21→26 0.00 29.76 0.00 92.56 0.00 11.92
Investigating the results in Tables (3)-(5) show the following findings

(i) The MPLS network with CR-LDP TE signal protocol has better performance in term of link
utilization. The RSVP reserves certain paths for transmission, while the CR-LDP utilizes most
14
of the available links. Table (6) shows the total number of links utilized by the two protocols
for the three networks.
Table (6): Total number of links utilized by the MPLS networks with two TE signals protocols
TE signal Total number of utilized links

Network
protocol Low load case Medium load case High load case
CR-LDP 15 15 16
Network 1
RSVP 10 11 14
CR-LDP 25 23 22
Network 2
RSVP 17 21 18
CR-LDP 22 24 28
Network 3
RSVP 14 15 20
(iii) The CR-LDP TE signal protocol manages almost equally to recognize t the size of the load
transmitted in both directions. Therefore, equal utilization of the links in both directions is
observed.
(iv) The CR-LDP capabilities for utilizing transmission links is affected by network topology. The
difference also can be noticed in the number of utilized links as given in Table (6).
(v) When Network 3 operates under high load condition, the RSVP TE signal protocol reservation
of certain links leads to congestions in some paths (path 14→22, path (22→14). However, it
uses these two paths at nearly full rate (93%) in the uplink which reveals the inability of this
protocol to manage the utilization between uplink traffic and downlink traffic on the same
path.
The investigation is carried further to address the ability of MPLS networks operating with CR-LDP and
RSVP TE signal protocols to detect and recover fault links. The results are given for Network 3 only since
other networks have similar behavior. Figure (7a) shows CR-LDP TE signal protocol capabilities to detect
transmission links failure and the speed of recovery. Figure (7b) presents RSVP TE signal protocol
capabilities to fault links detection and speed to recover it. Fail link time is considered at 100 sec
and recover time is set to 200 sec. The results show that both CR-LDP and RSVP detect the fail
links at the same time. For fault link recovery, the results show that the CR-LDP protocol is faster
than RSVP protocol in terms of recovered links. Unlike the RSVP protocol, the CR-LDP protocol
starts using the link immediately after the link being reconnected.
(5a) (5b)
Figure (7): Utilization of (node12→ node18) link. CR-LDP TE protocol (b) RSVP TE protocol

No. 5, October (2015); pp: 1-27
6 Links Utilization of QoS-Supported MPLS Networks
This section presents performance investigation of MPLS Traffic Engineering signal protocol capabilities
to utilize the transmission link under Quality of Service (QoS) conditions. The used QoS algorithms are
First-In First-Out (FIFO), Priority Queuing (PQ), Weighted Fair Queuing (WFQ) and Custom Queuing (CQ).
The results are represented for the three networks addressed in Section 5 with the same parameters and
applications. The average link utilization of the first network under different QoS algorithms is listed in
Tables (7)-(9) for low- , medium-, and high-load conditions, respectively. Results related to the second and
third networks are given in the Appendix.
Tables (10a)-(10c) list total number of links utilized by the two protocols in the three networks when
applying QoS for the three load conditions, respectively. These tables show that the four queuing
algorithms have similar effect (in term of utilized links) on MPLS networks operating with CR-LDP TE signal
protocol. This is not true when the RSVP TE signal protocol is used.
7 Conclusions
The performance of the MPLS CR-LDP and RSVP Traffic Engineering signal protocols has been evaluated
and compared in the term of link utilization with and without applying QoS. Different networks scenarios
and different applications, including video conferencing, voice, E-Mail, FTP, HTTP, DB, print and telnet
traffic have been.. The main conclusions drawn from this study are
(i) MPLS network with CR-LDP TE signal protocol has better performance in the term of link
utilization. The RSVP reserves certain paths for transmission while the CR-LDP utilizes almost
all of the available links.
(ii) The CR-LDP protocol is even better in terms of link management between uplink traffic and
downlink traffic on the same path.
(iii) The CR-LDP TE signal protocol is faster than RSVP protocol in terms of discovery of recovered
links, it starts using the link immediately after the link being reconnected.
(iv) There is inefficient link utilization with the RSVP protocol in which the reservation of specific
links for transmission produces transmission failure and packets drop at high loads.
(v) The MPLS TE signal protocol capability of link utilization is almost independent of network
topology or number of utilized links.
(vi) Applying QoS improves the performance of RSVP TE signal protocol while the number of
utilized links increases. For CR-LDP the same number of links is used before and after applying
QoS.
Table (7): Average link utilization of Network 1 operating with QoS under low load condition.
Utilization %
Link CR-LDP RSVP
FIFO PQ WFQ CQ FIFO PQ WFQ CQ
14→20 70.70 70.36 70.98 70.97 14.90 15.23 15.05 15.12
20 →14 1.72 27.00 31.00 30.79 0.86 0.80 1.06 1.04
18→15 47.98 54.23 45.98 46.01 69.42 69.42 56.71 56.74
15→18 75.37 70.96 70.35 70.33 0.87 0.87 1.28 1.29
16
18→21 49.18 45.12 49.20 49.27 0.33 0.34 0.74 0.75

21→18 0.85 9.79 1.21 1.31 69.36 69.36 54.44 54.42
19→15 25.72 26.88 27.92 27.97 12.52 12.63 12.65 12.63
15→19 15.014 1.57 14.19 13.98 69.73 69.98 69.96 69.95
19→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 →18 46.88 48.92 44.97 44.90 2.39 2.61 2.41 2.50
18→20 0.00 26.17 21.28 21.18 0.55 0.54 0.55 0.55
20→19 24.08 21.58 26.37 26.42 12.52 12.63 12.65 12.63
19→20 1.72 1.02 9.94 9.82 0.33 0.27 0.52 0.50
21→14 54.13 45.53 53.53 53.51 69.69 69.94 69.95 69.95
14→21 2.64 15.02 2.92 3.02 69.63 69.37 54.45 54.43
21→19 1.80 5.51 1.76 1.76 0.00 0.00 0.00 0.00
19→21 13.31 0.60 4.53 4.44 69.41 69.72 69.44 69.45
21→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→21 0.00 0.00 0.00 0.00l 0.00 0.00 0.00 0.00
Table (8): Average link utilization of Network 1 operating with QoS under medium load condition.
Utilization %
Link CR-LDP RSVP
70.70 70.97 71.02 70.36 17.13 17.38 66.34 66.32
14→20
20 →14 30.95 17.39 22.84 22.90 69.92 70.00 66.95 66.95
18→15 23.28 23.95 28.06 31.66 69.52 69.53 67.35 67.39
15→18 70.39 12.33 71.02 70.98 12.30 12.33 12.26 12.26
18→21 49.23 50.00 49.63 49.23 0.00 0.00 12.07 12.07
21→18 2.06 69.32 2.07 10.59 69.31 69.32 1.05 1.16
19→15 52.62 51.74 48.98 55.34 12.13 12.15 0.00 0.00
15→19 14.11 69.46 1.99 1.72 69.43 69.46 68.76 68.71
19→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 →18 21.37 21.90 26.19 21.27 5.06 5.32 66.34 66.32
18→20 21.30 12.33 21.75 22.09 12.30 12.33 0.20 0.20
20→19 49.58 49.41 45.19 49.22 12.07 12.07 0.00 0.00
19→20 9.86 68.44 1.28 1.00 68.93 68.44 66.77 66.77
21→14 53.55 50.61 50.17 49.79 0.50 1.02 14.06 14.02
14→21 6.25 69.40 6.02 18.28 69.37 69.40 1.14 1.25
21→19 4.24 3.71 40.00 7.97 0.00 0.00 0.00 0.00
19→21 4.52 1.02 0.74 0.75 0.50 1.02 2.00 1.95
21→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→21 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

No. 5, October (2015); pp: 1-27
Table (9): Average link utilization of Network 1 operating with QoS under high load condition.
Utilization %
Link CR-LDP RSVP
14→20 70.45 70.45 70.50 71.05 70.02 28.98 27.20 27.48
20 →14 61.84 61.0 62.46 61.20 69.63 69.98 18.03 17.59
18→15 60.82 61.30 62.38 70.39 1.03 69.53 17.29 17.68
15→18 70.39 70.39 70.43 70.39 0.98 69.41 65.92 65.60
18→21 21.26 21.47 21.63 7.07 0.98 1.07 65.91 65.59
21→18 15.96 16.56 15.86 7.07 0.00 68.41 0.00 0.00
19→15 36.83 34.77 36.19 36.74 69.90 21.86 69.56 69.46
15→19 30.24 28.81 29.54 29.41 70.40 19.87 20.40 19.85
19→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20 →18 48.51 48.72 48.20 44.42 1.03 8.11 17.29 17.68
18→20 49.27 49.07 48.96 49.31 0.00 68.35 0.00 0.00
20→19 22.08 21.88 22.47 26.99 69.00 20.88 9.92 9.81
19→20 14.39 17.89 14.19 18.32 69.63 6.99 18.03 17.59
21→14 37.18 32.50 37.07 32.39 1.75 13.95 67.73 67.43
14→21 30.63 29.36 29.52 16.96 2.23 69.39 65.87 65.51
21→19 15.00 13.13 13.99 9.98 2.23 0.98 65.87 65.51
19→21 16.19 11.27 15.69 11.42 0.77 12.89 2.38 2.27
21→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→21 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Table (10a): Total number of links utilized by the two protocols when applying QoS (low load condition).

Network
protocol FIFO PQ WFQ CQ
CR-LDP 15 16 16 16
Network 1
RSVP 15 15 15 15
CR-LDP 18 18 18 18
Network 2
RSVP 15 16 16 16
CR-LDP 20 19 20 19
Network 3
RSVP 25 23 23 239
Table (10b): Total number of link utilized by the two protocols when applying QoS (medium load condition).

Network
CR-LDP 16 16 16 16
Network 1
RSVP 14 14 13 13
CR-LDP 15 15 15 15
Network 2
RSVP 20 21 20 20
CR-LDP 25 25 25 25
Network 3
RSVP 22 23 24 22
18
Table (10c): Total number of link utilized by the two protocol when applying QoS (high load condition).
Network
CR-LDP 16 16 16 16
Network 1
RSVP 15 15 14 15
CR-LDP 18 18 18 18
Network 2
RSVP 20 17 1 19
CR-LDP 22 22 22 22
Network 3
RSVP 21 18 20 20
REFERENCES
[1] A. Hussain, S. Nazeer, T. Abdullah, B. Yaseen and A. Salam “Network Resilience in Multiprotocol
Label Switching” Journal of Asian Scientific Research, Vol.2, No.4, PP. 221-227, April 2012.
[2] B. Forouzan, “Data Communications and Networking”, McGraw-Hill Higher Education, 4th Ed.,
PP. 101-110, 2006.
[3] Dr. R. Naoum and M. Maswady, ”Performance Evaluation for VoIP over IP and MPLS”, World of
Computer Science and Information Technology Journal, Vol. 2, No. 3, PP. 110-114, May 2012.
[4] T.J. Shi, G. Mohan, “An Efficient Traffic Engineering Approach Based on Flow Distribution and
Splitting in MPLS Networks”, Computer Communications, Vol. 29, No. 9, PP. 1284-1291, May
2006.
[5] A. Bongale, ”Analysis of Link Utilization in MPLS Enabled Network using OPNET IT Guru”,
International Journal of Computer Applications, Vol. 41, No.14, PP. 35-40, March 2012.
[6] H. Perros, “An Introduction to ATM Network”, 1st Ed., Cary, USA, 2001.
[7] R. Peterkin “A Reconfigurable Hardware Architecture for VPN MPLS based Services”, MSc.
Thesis, School of Information Technology and Engineering, University of Ottawa, April 2006.
[8] Y. Lee, Y. Seok, Y. Choi and C. Kim “A Constrained Multipath Traffic Engineering Scheme for
MPLS Networks”, Electronic Telecommunication Research Institute, 2002.
[9] S. Shyry and V. Ramachandran, “Finagling Congestion in Selfish Overlay Routing Belittling Link
Utilization” , Proc. IEEE, PP. 308-311, November 2010.
[10] C. Pelsser , O. Bonaventure “Extending RSVP-TE to support Inter-AS LSPs”, Proc. CiteSerr,Vol
41, No.5, PP. 122-128, May 2011.
[11] T. Aziz, N. Islam khan and A. Popescu ,“Effect of Packet Delay Variation on Video/Voice over
DIFFSERV-MPLS in IPV4/IPV6 Networks”, International Journal of Distributed and Parallel
Systems, Vol.3, No.1, pp.27-47, January 2012.

No. 5, October (2015); pp: 1-27
[12] M. Bhandure, G. Deshmukh and Prof. V. J n, “ Comparative Analysis of Mpls and Non -Mpls
Network”, International Journal of Engineering Research and Applications, Vol. 3, No. 4, PP.
71-76, August 2013.
[13] S. Ibrahim, ” Performance Evaluation of Traffic Engineering Signal Protocols in MPLS Network “,
M.Sc. Thesis, Al-Nahrain University, April 2013.
[14] A. Sulaiman and O. Alhafidh“ Performance Analysis of Multimedia Traffic over MPLS
Communication Networks with Traffic Engineering“, International Journal of Computer
Networks and Communications Securitik Vol. 2, No. 3, PP. 93 -101, March 2014.
[15] S. Ibrahim and M. AL-Quzwini, “Performance Evaluation of MPLS TE Signal Protocols with
Different Audio Codecs for Voice Application”, International Journal of Computer Applications,
Vol. 57, No.1, pp. 56-60, November 2012.
[16] H. Perros, “An Introduction to ATM Network”, 1st Ed., Cary, USA, 2001.
[17] R. Peterkin “A Reconfigurable Hardware Architecture for VPN MPLS based Services”, MSc.
Thesis, School of Information Technology and Engineering, University of Ottawa, April 2006.
[18] T. Latif and K. Malkajgiri, “Adoption of Voice over Internet Protocol,” Lulea University of
Technology, 2007.
[19] O. Dokun and A. Gift “PDH (Plesiochronous Digital Hierarchy) /SDH-SONET (Synchronous Digital
Hierarchy / Synchronous Optical Networking)”, International Journal of Mathematics and
Engineering Research, Vol 3, No 1, January 2015.
[20] A. Kumar and S. G. Thorenoor, “Analysis of IP Network for Different Quality of Service.” In
International Symposium on Computing, Communication, and Control (ISCCC), Proc .of CSIT Vol.
1. IACSIT Press, Singapore. 2011.
[21] P. Ivaniš and D. Drajic “The Simulation Model of Optical Transport System and Its Applications
to Efficient Error Control Techniques Design”, Electronic, Vol. 13, No. 2, December 2009.
[22] K. Januu and R. Deekona,“ OPNET Simulation of Voice Over MPLS With Considering Traffic
Engineering”, M.Sc. Thesis, School of Engineering, Blekinge Institute of Technology, Sweden,
June 2010.
[23] J. Evans and C. Filsfils,”Deploying IP and MPLS QoS for Multiservice Networks”, 1st Ed., Elsevier
Inc, San Francisco, USA, 2007.
[24] V. Alwayn, “Advanced MPLS design and Implementation, Cisco Systems”, Cisco press 201, West
103rd Street Indianapolis, 2001.
20
[25] K. Salah and A. Alkhoraidly “An OPNET-Based Simulation Approach for Deploying VoIP”,
International Journal of Network Management Int. J. Network, DOI: 10.1002, PP. 159–183, 27
January, 2006.
[26] H. Asif and G. Kaosa, “Performance Comparison of IP, MPLS and ATM Based Network Cores
using OPNET”, IEEE Computer Society, August, 2006.
[27] A. Hadi, “Modeling Computer Laboratories in University that Contains Eight Colleges by Using
OPNET Software”, Journal of Kerbala University, Vol. 7, No.1, PP.108-123, 2009.
[28] S. Kuribayashi,” Proposed Optimal LSP Selection Method in MPLS Networks”, International
Journal of Computer Networks & Communications Vol.3, No.1, PP. 241-250, January 2011

No. 5, October (2015); pp: 1-27
Appendix: Effect of QoS on the Performance of Networks 2 and 3

Table (A1): Average link utilization of Network 2 operating with QoS (low load).
Utilization %
Link CR-LDP RSVP
14→23 69.78 70.21 70.10 70.21 69.89 70.1 66.09 65.98
23→14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
25→20 0.22 0.40 1.05 0.31 0.00 0.00 0.00 0.00
20→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26 →20 69.73 70.37 70.00 70.15 13.04 12.72 14.07 14.03
21→19 0.70 0.83 0.78 0.71 0.00 0.00 0.00 0.00
19→21 0.15 0.23 0.91 0.21 69.48 69.49 64.55 64.55
21→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→21 69.76 70.40 70.05 70.18 0.00 12.72 14.07 14.07
21→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→21 0.00 0.00 0.00 0.00 2.89 0.00 0.00 0.00
22→14 69.56 69.89 69.76 69.89 70.08 70.16 70.06 70.06
14→22 1.77 2.03 1.91 2.12 70.08 2.29 2.58 2.55
22→21 0.70 0.85 0.79 0.72 0.00 0.00 0.00 0.00
21→22 69.78 70.42 70.09 70.20 70.08 70.16 70.06 70.06
22→24 1.10 1.17 1.15 1.43 2.90 2.29 2.59 2.55
24→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→24 69.77 70.20 70.09 70.20 69.89 70.15 66.08 65.98
24→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→19 69.78 70.21 70.10 70.20 69.91 70.18 67.92 67.87
19→24 0.00 0.00 0.00 0.00 69.49 69.50 64.55 0.00
25→15 69.53 69.85 69.74 69.86 69.90 70.18 67.92 67.87
15→25 0.37 0.63 1.94 0.53 69.90 69.50 67.92 64.55
25→19 0.19 0.27 0.95 0.25 69.49 69.50 64.55 64.55
19→25 69.78 70.41 70.10 70.20 13.04 12.72 14.07 14.03
26→15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
15→26 69.73 70.37 70.00 70.15 0.00 0.00 0.00 0.00
26→25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
25→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22
Table (A2): Average link utilization of Network 2 operating with QoS (medium load).
Utilization %
Link FIFO PQ WFQ CQ FIFO PQ WFQ CQ
CR-LDP CR-LDP CR-LDP CR-LDP RSVP RSVP RSVP RSVP
14→23 69.90 70.16 70.21 70.17 1.66 1.47 14.10 14.05
23→14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
19→20 46.14 0.00 0.00 0.00 0.00 69.50 0.00 0.00
20→25 0.00 0.00 0.00 0.00 0.00 57.53 0.00 0.00
25→20 43.22 43.99 44.20 43.40 7.00 14.16 1.30 1.28
20→26 0.00 0.00 0.00 0.00 0.24 0.12 1.10 1.04
26 →20 0.27 0.22 1.11 0.31 2.10 0.00 70.08 70.15
21→19 0.00 0.00 0.00 0.00 63.03 0.38 3.05 2.92
19→21 23.88 26.39 26.21 26.70 9.10 70.18 0.00 0.00
21→20 0.00 0.00 0.00 0.00 0.24 57.64 1.10 1.03
20→21 46.40 44.19 45.30 43.70 0.00 0.00 70.12 70.19
21→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→21 0.00 0.00 0.00 0.00 70.10 70.21 0.00 0.00
22→14 69.57 69.77 69.77 69.79 70.07 70.17 70.12 70.11
14→22 0.00 0.00 0.11 0.00 70.07 70.18 56.88 56.68
22→21 0.00 0.00 0.00 0.00 63.28 58.02 4.14 3.95
21→22 69.80 70.07 70.10 69.84 0.00 0.00 70.12 70.19
22→24 0.00 0.00 0.00 0.50 6.82 12.19 52.74 53.67
24→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→24 69.90 70.16 70.20 70.16 1.66 1.47 14.10 14.05
24→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→19 69.90 70.16 70.20 70.16 8.47 13.65 66.068 66.26
19→24 0.00 0.00 0.00 0.50 70.03 69.50 0.00 0.00
25→15 69.64 69.81 69.84 69.82 70.13 70.26 68.92 68.78
15→25 69.78 70.07 70.08 70.22 70.14 14.03 1.30 1.28
25→19 23.90 26.43 26.23 26.86 63.03 0.00 0.00 0.00
19→25 69.90 70.17 70.21 70.31 2.10 14.03 68.92 68.88
26→15 0.00 0.00 0.00 0.00 0.24 0.12 1.10 1.04
15→26 0.27 0.21 1.08 0.30 0.00 0.00 70.09 70.15
26→25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
25→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

No. 5, October (2015); pp: 1-27
Table (A3): Average link utilization of Network 2 operating with QoS (high load)
Utilization %
Link CR-LDP RSVP
14→23 60.07 10.12 8.32 69.49 28.33 15.51 35.20 35.71
23→14 0.00 0.00 0.00 0.00 14.28 0.38 69.82 69.66
20→19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.13
19→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.69
20→25 0.00 0.00 0.00 0.12 69.48 0.00 0.00 0.00
25→20 49.30 49.38 44.20 10.20 0.40 0.00 0.40 0.40
20→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26 →20 0.22 0.59 1.11 69.25 69.93 69.80 17.66 16.02
21→19 21.47 21.55 21.68 6.47 0.00 0.00 53.99 53.87
19→21 21.25 21.64 26.21 4.54 0.00 0.00 0.00 0.00
21→20 0.00 0.00 0.20 0.19 69.48 0.00 0.00 0.00
20→21 49.50 49.91 45.30 70.47 69.97 69.51 18.05 53.87
21→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→21 0.00 0.00 0.00 0.00 0.00 69.80 0.00 16.41
22→14 70.05 70.34 70.33 70.36 69.48 70.41 18.05 16.42
14→22 69.93 70.20 0.11 18.20 69.63 12.14 54.00 0.00
22→21 21.50 21.68 21.88 6.65 69.48 0.00 54.00 53.87
21→22 70.31 70.28 70.10 6.65 69.97 0.00 18.05 0.00
22→24 48.70 48.88 48.71 11.83 0.00 69.51 0.00 0.00
24→22 0.00 0.00 0.00 0.00 0.00 0.38 0.00 0.00
23→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22→23 0.00 0.00 0.00 0.00 0.00 12.51 0.00 0.00
23→24 6.11 10.17 8.37 69.49 28.33 15.51 35.20 35.71
24→23 0.00 0.00 0.00 0.00 14.29 12.60 69.82 69.82
24→19 53.75 57.14 57.04 70.66 28.33 70.81 35.20 35.71
19→24 0.00 0.00 0.00 0.00 14.29 70.81 69.82 69.82
25→15 70.06 70.16 70.23 70.15 0.00 70.81 70.44 70.26
15→25 70.28 70.20 70.08 14.61 14.29 69.80 70.22 70.22
25→19 21.26 21.65 21.62 4.68 14.23 12.51 69.83 69.22
19→25 70.26 70.36 70.21 70.37 28.33 0.00 70.45 70.22
26→15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
15→26 0.21 0.55 1.09 69.25 69.48 0.00 17.65 16.02
26→25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.0
25→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24
Table (A4): Average link utilization of Network 3 operating with QoS (low load)
Utilization %
LINK RSVP CRLDP
FIFO PRI WFQ CUST FIFO PRI WFQ CUST
12→18 70.52 71.15 55.22 44.40 12.78 12.73 13.44 0.64
18→12 0.52 0.54 0.53 0.55 0.63 12.79 0.63 0.63
12→20 0.54 0.54 0.53 12.70 0.19 0.20 0.23 12.40
20→12 0.60 0.69 0.63 0.62 48.12 48..01 47.85 43.48
12→21 0.11 0.17 12.30 0.15 70.25 70.19 70.22 70.18
21→12 0.57 12.72 0.57 0.57 12.17 0.00 12.17 0.00
12→24 0.52 0.54 0.53 0.54 0.63 12.79 0.63 0.63
24→12 70.52 71.15 55.21 44.39 12.76 12.73 13.44 0.64
14→17 70.89 70.54 66.83 66.76 0.00 0.00 0.00 0.00
17→14 12.06 0.00 0.00 0.00 0.00 0.00 0.47 0.13
14→22 0.60 0.69 0.63 0.62 70.29 70.22 70.29 70.84
22→14 1.04 1.04 1.04 13.21 0.21 0.20 0.55 12.43
14→23 0.52 0.54 0.53 0.54 0.65 13.04 0.65 0.65
23→14 70.52 71.15 55.22 44.39 13.02 12.98 13.70 0.65
14→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→18 70.88 70.53 66.82 66.75 22.30 22.21 22.45 27.37
18→17 12.56 0.50 0.52 0.51 0.00 0.00 0.79 0.15
18→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
21→20 12.06 0.00 0.00 0.00 0.00 0.00 0.48 0.13
20→21 70.88 70.54 66.83 66.75 0.00 0.00 0.00 0.00
21→23 0.11 0.17 12.30 0.15 70.12 70.06 70.08 70.06
23→21 0.57 12.72 0.57 0.57 12.17 0.00 12.17 0.00
22→17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22→24 0.57 12.72 0.57 0.57 12.42 0.00 12.42 0.00
24→22 0.11 0.16 12.30 0.15 70.18 70.18 70.22 70.18
23→17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→20 0.60 0.69 0.63 0.62 48.05 48.08 47.93 43.66
20→26 0.54 0.54 0.53 12.70 0.21 0.22 0.25 12.66
26→21 0.00 0.00 0.00 0.00 22.37 22.28 22.53 27.45
21→26 0.51 0.50 0.52 0.51 0.00 0.00 0.00 0.00

No. 5, October (2015); pp: 1-27
Table (A5): Average link utilization of Network 3 operating with QoS (medium load).
Utilization %
LINK RSVP CR-LDP
12→18 1.09 0.19 1.18 13.31 1.11 1.15 1.106 13.23
18→12 14.1 1.97 1.82 1.82 1.08 13.16 1.02 13.17
12→20 0.00 0.00 12.17 0.00 1.00 1.00 1.00 1.03
20→12 0.11 0.12 0.00 0.00 2.56 2.56 2.34 2.74
12→21 13.17 0.48 1.02 1.02 12.22 12.32 12.30 0.17
21→12 3.83 4.07 4.20 3.83 70.31 70.06 70.46 70.06
12→24 14.09 1.97 1.82 1.82 1.08 13.16 1.02 13.17
24→12 1.09 0.19 1.18 13.31 1.11 1.15 1.11 13.23
14→17 70.55 70.51 55.37 66.79 0.00 0.00 0.00 0.00
17→14 68.89 57.73 43.67 43.65 22.12 22.09 21.97 22.25
14→22 1.13 13.37 13.22 1.12 7.05 7.22 6.94 7.40
22→14 1.69 13.43 14.09 1.77 49.13 49.12 49.32 49.00
14→23 14.09 1.97 1.82 1.82 1.10 13.42 1.04 13.43
23→14 1.09 0.19 1.18 13.31 1.12 1.17 1.12 13.49
14→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→18 70.57 71.13 68.00 67.78 4.58 4.75 4.70 4.76
18→17 70.57 71.17 45.57 45.41 70.27 70.24 70.29 70.23
18→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
21→20 68.89 57.74 43.67 43.65 22.19 22.16 22.05 22.31
20→21 70.54 4.07 55.36 66.79 0.11 0.11 0.11 0.11
21→23 13.17 0.48 1.02 1.02 12.22 12.32 12.30 0.17
23→21 3.83 0.00 4.21 3.83 70.56 70.19 70.81 70.18
22→17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→22 0.00 0.48 0.00 0.00 0.00 0.00 0.00 0.00
22→24 3.83 4.06 4.21 3.83 70.57 70.18 70.80 70.18
24→22 13.17 0.00 1.02 1.02 12.47 12.57 12.55 0.18
23→17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→20 0.11 0.12 0.00 0.00 2.69 2.58 2.37 2.76
20→26 0.00 0.00 12.17 0.00 1.00 1.00 1.02 1.04
26→21 1.02 13.26 13.13 1.03 4.52 4.69 4.62 4.70
21→26 1.69 13.43 1.93 1.77 48.22 48.22 48.40 48.06
26
Table (A6): Average link utilization of Network 3 operating with QoS (high load).
Utilization %
LINK RSVP CR-LDP
FIFO PRI WFQ CUST FIFO PRI WFQ CUST
12→18 12.05 0.00 0.00 0.00 0.32 0.62 2.31 12.43
18→12 13.38 27.23 19.02 31.33 0.00 0.00 0.00 0.00
12→20 0.00 0.00 0.00 12.17 0.00 0.00 0.00 0.00
20→12 6.92 7.71 10.77 10.13 2.90 13.10 5.55 3.62
12→21 2.41 2.15 2.33 2.40 0.46 1.36 2.44 .68
21→12 0.00 0.00 0.00 0.00 70.10 70.11 70.16 70.50
12→24 13.38 27.23 19.03 31.34 12.10 0.00 0 0
24→12 12.05 0.00 0.00 0.00 0.32 0.61 2.31 12.43
14→17 70.84 70.51 54.63 54.64 10.00 10.00 22.01 10.05
17→14 1.50 13.68 42.43 42.40 24.84 23.03 22.32 23.19
14→22 9.33 1.00 25.24 12.33 7.40 25.90 14.01 11.89
22→14 69.25 57.69 20.76 24.32 46.18 48.28 49.01 47.50
14→23 13.38 27.23 19.02 31.33 0.00 0.00 0.00 0.00
23→14 12.05 0.00 0.00 0.00 0.33 0.62 2.33 12.68
14→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→18 70.84 70.54 68.1 56.67 14.51 22.66 30.56 18.31
18→17 70.75 71.37 63.19 54.56 70.80 71.08 71.11 70.45
18→20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
20→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
21→20 1.5 13.68 42.43 42.41 24.97 23.65 22.50 23.26
20→21 70.84 70.51 54.63 54.63 10.04 9.96 22.27 10.08
21→23 2.41 2.15 2.33 2.40 0.46 1.36 2.44 0.68
23→21 0.00 0.00 0.00 0.00 70.23 70.26 70.33 70.34
17→22 0.00 0.00 0.00 0.00 00.00 0.00 0.00 0.00
22→24 0.00 0.00 0.00 0.00 70.23 70.38 70.33 70.85
24→22 2.41 2.15 2.33 2.40 0. 47 1.36 2.46 0.69
23→17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
17→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
23→26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
24→18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18→24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
26→20 6.92 7.71 10.77 10.13 2.92 13.38 5.57 3.65
20→26 0.00 0.00 0.00 12.16 0.25 0.25 0.25 0.25
26→21 2.43 2.30 14.48 2.21 4.52 12.82 8.58 8.29
21→26 69.26 57.70 20.76 12.16 46.1 48.24 48.99 47.34

211 MPLS Te

Uploaded by

Copyright:

Available Formats

211 MPLS Te

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

211 MPLS Te

Uploaded by

Copyright:

Available Formats

SOCIETY FOR SCIENCE AND EDUCATION

Links Utilization in MPLS Networks Operating with Traffic

Copyright © Society for Science and Education United Kingdom 3

3.1 MPLS Network

Figure (2): MPLS domain network

Figure (3): MPLS planes [17]

3.2 Traffic Engineering Signal Protocols

Copyright © Society for Science and Education United Kingdom 5

3.3 Constraint-Based Label Distribution Protocol

3.4 Resource Reservation Protocol

4 Simulated Network Topologies and Setup Parameters

4.1 Network Topologies

Copyright © Society for Science and Education United Kingdom 7

4.2 Parameters used in the Simulation

Table (2): Applications parameters

Copyright © Society for Science and Education United Kingdom 9

5 Link Utilization in the Absence of QoS

Copyright © Society for Science and Education United Kingdom 11

Table (4): Average link utilization of Network 2 in the absence of QoS.

Copyright © Society for Science and Education United Kingdom 13

Investigating the results in Tables (3)-(5) show the following findings

TE signal Total number of utilized links

Copyright © Society for Science and Education United Kingdom 15

18→21 49.18 45.12 49.20 49.27 0.33 0.34 0.74 0.75

Copyright © Society for Science and Education United Kingdom 17

TE signal Total number of utilized links

TE signal Total number of utilized links

Copyright © Society for Science and Education United Kingdom 19

Copyright © Society for Science and Education United Kingdom 21

Appendix: Effect of QoS on the Performance of Networks 2 and 3

Copyright © Society for Science and Education United Kingdom 23

Copyright © Society for Science and Education United Kingdom 25

Copyright © Society for Science and Education United Kingdom 27

You might also like