Ad Hoc Networks 17 (2014) 114128

Contents lists available at ScienceDirect

Ad Hoc Networks
journal homepage: www.elsevier.com/locate/adhoc

A MAC protocol with mobility support in cognitive radio ad hoc

networks: Protocol design and analysis
Peng Hu, Mohamed Ibnkahla
Department of Electrical and Computer Engineering, Queens University, Kingston, ON K7L 3N6, Canada

a r t i c l e

i n f o

Article history:
Received 28 July 2012
Received in revised form 8 November 2013
Accepted 18 January 2014
Available online 24 January 2014
Keywords:
Mobility
Medium access control
Throughput analysis
Cognitive radio ad hoc networks

a b s t r a c t
Cognitive radio ad hoc networks (CRAHNs) have recently been proposed as a way to bring
cognitive radio (CR) technology to traditional ad hoc networks. An important problem is to
design a medium access control (MAC) protocol that addresses the decentralized control
and local observation for spectrum management. In this paper, we propose a cognitive
MAC protocol with mobility support (CM-MAC) based on carrier sense multiple access/collision avoidance (CSMA/CA) technique, where our protocol can respond to the vicinity state
of CR nodes to primary exclusive regions. This paper analyzes the throughput performance
for the proposed MAC protocol with the consideration of multiple PU activities and CR
nodes. Our analytical results show that the proposed MAC protocol outperforms the
throughput performance of the classical CSMA/CA MAC and statistical channel allocation
(SCA) MAC protocols.
2014 Elsevier B.V. All rights reserved.

1. Introduction
As a result of the development of cognitive radio technology, the concept of cognitive radio ad hoc networks
(CRAHNs) has recently been proposed in the literature
[1], which involve more challenges than those in classical
cognitive radio networks (CRNs). These challenges are
due to variable radio environments caused by spectrumdependent communication links, hop-by-hop transmission, changing topology, and node mobility.
Different from traditional medium access control (MAC)
protocol used in ad hoc networks, the MAC protocol for
CRAHNs has to address the spectrum sharing function [2],
as well as to improve the throughput and spectral efciency.
Furthermore, because the primary exclusive region (PER) [3]
of primary users (PUs) is an important factor that can make a
signicant impact on CR and PU communications, a scheme
addressing the PER should be considered in MAC protocols.
Corresponding author. Tel.: +1 613 533 3074; fax: +1 613 533 6615.
E-mail addresses: peng.hu@queensu.ca (P. Hu), mohamed.ibnkahla@
queensu.ca (M. Ibnkahla).
http://dx.doi.org/10.1016/j.adhoc.2014.01.008
1570-8705/ 2014 Elsevier B.V. All rights reserved.

Furthermore, the classical carrier sense multiple access/

collision avoidance (CSMA/CA) based MAC protocols have
the advantage of solving hidden terminal problems and
having distributed operations (e.g., distributed coordination function in IEEE 802.11 MAC); thus, state-of-the-art
MAC protocols [49] for CRNs have been proposed. However, PER, PU/CR activity and PER have not been comprehensively addressed in the literature.
In this paper, we propose a CSMA/CA-based MAC protocol called CM-MAC for CRAHNs to improve the performance
of the network. Although the throughput performance can
be related to the poor network organization and routing as
the cross-layer nature of CRAHNs, the proposed mobility
support algorithm (MSA) solution to PER issues working
mostly at the MAC layer is mainly twofold. One is the MAC
layer is a right place to address the PER issues. If an upperlayer scheme like a routing scheme is employed, a MAC layer
mechanism is still needed to address the issues caused by
PER like spectrum sharing, mobility, and detection. The
other is solving the PER issues in MAC layer is lightweight
as either the path formation or scheduling in routing is

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

costly. For the topology, it might be an issue but we assume

the nodes have sufcient communication with each other in
the network. In summary, the main contributions of the paper are as follows:
1. We focus on a CM-MAC protocol that addresses CR
mobility and PER issues.
2. We analyze the throughput of CM-MAC protocol
assuming that the PU trafc follows a Poisson
process.
3. We show that the throughput and spectrum utilization are improved by CM-MAC compared to classical MAC protocols.
The remainder of the paper is organized as follows: Section 2 discusses the related work regarding MAC protocols
for CRAHNs; Section 3 presents the system model and
motivation; Section 4 describes the CM-MAC protocol; In
Section 5, the throughput analysis of the CM-MAC protocol
is given; Section 6 presents numerical results; and Section 7 concludes the paper and presents some future work.

2. Related work
The objectives of the CRAHN MAC protocol not only include the improvement of channel utilization and throughput without degrading PU communications, but also
include the control of spectrum management modules
such as spectrum access and spectrum sharing functions
to determine the timing for data transmissions [1].
The use of multiple channels for throughput improvement has been addressed in several MAC protocols. A feasible solution for throughput improvement is to nd a set of
good quality channels. A dual-channel MAC protocol
(DUCHA) was proposed in [10] which can improve the
one-hop throughput up to 1.2 times and multi-hop
throughput up to ve times compared to the IEEE 802.11
MAC protocol. An opportunistic multi-radio MAC (OMMAC)
was proposed in [11], where a multi-channel-based packet
scheduling algorithm was employed and packets were sent
on a channel having best spectral efciency (i.e., the channel with the highest bit rate).
In a CRN, the spectrum utilization can be improved if
we choose the appropriate set of channels that meet the
transmission rate requirement. A MAC protocol based on
statistical channel allocation (SCA) was proposed in [4]
which uses a channel aggregation approach to improve
the throughput and dynamic operating range to reduce
the computational complexity. Results of [4] show that
SCA-MAC can use spectrum holes effectively to improve
spectrum efciency while keeping the performance of
coexisting PUs. However, the PER issue was not addressed
explicitly in [4]. In order to meet data rate requirement for
data transmissions, a MAC with a so-called multi-channel
parallel transmission protocol was proposed in [7], where
the minimum number of channels was selected to meet a
certain data rate. The results of [7] show that the proposed
MAC protocol improves spectrum utilization and system
throughput than the results shown in [8], where the
proposed scheme selects a channel by the best signal-to-

115

interference-plus-noise ratio (SINR) value. In [9], an opportunistic auto-rate MAC protocol was used to maximize the
utilization on individual channels.
Spectrum sharing and spectrum access functions are
explicitly addressed in [6], where spectrum access and
spectrum allocation schemes are introduced in the proposed cognitive radio MAC (COMAC) protocol. Specically,
the spectrum utilization is improved by providing enough
channels instead of assigning all the possible channels to a
CR node, so that the other available channels could be reserved for other CR transmissions. In [5], the authors employed a distance-dependent channel assignment scheme
in a proposed distance-dependent MAC (DDMAC).
However, the aforementioned works did not comprehensively consider several important factors. Firstly,
although the spectrum sensing can be simultaneously performed in one shot [12], the sensing time cannot be ignored, as it may be relatively long, which may lead to
end-to-end throughput degradation [13]. Secondly, with
the existence of PER where CR communications tend to
interfere with PU communications, the CR should retain silent when moving into this region if maintaining PU communication is a priority.
Because CRAHN MAC protocols favor distributed solutions, the distributed function like distributed coordination
function (DCF) is a good option for protocol design. In fact,
most of the aforementioned MAC protocols [59,11] are
DCF-based with request-to-send (RTS)/clear-to-send (CTS)
handshaking procedures, which intrinsically solved the hidden terminal problem. Other non-CSMA/CA-based MAC
protocols like multi-channel MAC (MMAC) [14] and cognitive MAC (C-MAC) [15] can also solve the hidden terminal
problem; however, they require a periodic synchronization
which can hardly be applied to large-scale CRAHNs.

3. System model and motivation

3.1. System model
Before further discussion, we describe the system model used in this paper. A CRAHN is deployed in a plane containing Np PUs and NCR CR nodes. In a certain time period, a
set of channels, denoted by Ki(t), is available to a CR node i
and thus the total number of channels available to CR node
i is |Ki(t)|. The set of channels on a transmission link between ith CR and (i + 1)th CR is Ki,i+1. There are K spectrum
bands in total available to CRs and PUs, while the typically
used (K + 1)th out-of-band common control channel (CCC)
[16] is employed to exchange control information. When
the jth PU is active for a transmission, its trafc ow takes
one channel Ck, i.e., KPj t fC k g. For simplicity, we use the
following notation: KPj t fkg. When the jth PU is active,
KPj t fkjk > 1 and k  Kg. In the paper, the PU trafc
ow is assumed to follow a Poisson process with parameter k [17]. Note that in this paper, a PU that occupies multiple channels is equivalent to multiple PUs that occupy
different channels.
We consider a CRAHN with a PER as shown in Fig. 1,
where it is located at the center of the network and the primary receivers are within the PER bounded by the circle of

116

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

Fig. 1. A CRAHN with a PER and multiple CRs.

radius R0 + e and e is a small guard band around the PER. A

PER operating on channel k is denoted by SPER(k). SPER(k)
(i.e., the shaded area shown in Fig. 1) has a radius of
R0 + e and the radio coverage of a PU (i.e., the area circled
by dotted line in Fig. 1) has a radius of R, where R0 < R. In
the SPER(k), CR communications will severely affect the
PU communication, and vice versa.
In [18], the combination of two spectrum sharing models (a cooperation model and a coexistence model) with
two types of spectrum-sharing arrangements (i.e., sharing
among equals and primarysecondary sharing) was introduced. In this paper, we will adopt the spectrum sharing
combination of the coexistence-based model and primarysecondary sharing arrangement for CRAHNs. The
coexistence-based model means that devices try to avoid
interference without explicit signaling [18]; and the primarysecondary sharing arrangement results in the primary system having exclusive rights to access spectrum
bands through licensing.

3.2. Motivation
A MAC protocol needs to decide the availability of spectrum bands for current and future data transmissions.

These spectrum bands can facilitate the upper-layer protocol (e.g., routing protocol) in order to obtain an optimized
path for data transmissions. Moreover, in order to keep a
desirable throughput, a MAC protocol is expected to perform local observation without extra communication efforts. As such, the information exchange based on the
built-in handshaking procedure in a CSMA/CA MAC protocol is a preferred solution.
Traditionally, the MAC sub-layer is at the link layer,
where the link layer is in charge of the communication
between adjacent nodes. Therefore, based on the layered
perspective for the CRAHN protocol stack, maintaining
the communications with adjacent nodes while sharing
the spectrum resources among nodes is a major
challenge.
In order to determine why a CRAHN MAC is important for
data transmissions, we consider an example of a typical data
transmission scenario shown in Fig. 2, where node S tries to
transmit data frames to node E through the path from node
A to D. In the previous time slots, the data was transmitted
on channel 3, which was not occupied by either PU 1 or PU
2. However, channel 3 is now occupied by PU 2, and the link
from node C to D and the link from node D to E are broken.
Therefore, the rest of nodes need to be informed of the spectrum change of node D and E. Due to the nature of changing
PU activity, it is efcient to perform updates of spectrum
changes before each data transmission. Moreover, because
this PU spectrum availability is dependent on the PER region
induced by a PU, PER regions should be considered by a
CRAHN MAC protocol.

3.3. Discussion on PER region and throughput

To see how PER region can affect CR and PU throughput,
we consider the network shown in Fig. 1, where CRs are
distributed outside PER when CRs and PUs are operating
on channel k. In this sense, Mai et al. [3] derive the
worst-case interference power that the PU transmitter
experienced from all CR nodes where a PU transmitter
communicates with another PU receiver at a distance of
R0. This worst-case interference power is:

{1,2,3,5}
A

Channel 4 is not available

in this area

S
PU 1

Channel 3 is not available

in this area

{1,2,4,5}
D
E
PU 2

Fig. 2. An example of the necessity of a CRAHN MAC protocol in a CRAHN. The available spectrum bands for the nodes covered by a PU are shown in
brackets. The links are broken (shown in dashed arrows) when the data transmission from S to E is operated on channel 3.

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

R e2

R2

117

6
7
EI0 a4 dpP 4
5
2 2
e 2R0 e2
R2  R20

where a is the path loss exponent, R0 is the radius of SPER(k)

and R is the coverage radius of the PU; e is the guard band
radius, which ensures that the interference caused by CRs
will not affect PU communications. CR nodes are distributed in the outer circular area of PER with density h.
Because a CSMA/CA MAC protocol is based on timeframes, we consider a timeframe with an interval [0, T]. If
the PU transmitter/receiver pair is active for m time units,
while CR nodes are active for the entire timeframe, CRs
can interfere with PU communications in the m time units.
Based on the data rate equation in [3], and, in this case, the
data rate of PU, DPU, and data rate of CR, DCR, can be expressed as:

DPU 

DCR 6

m
T

m
T

log 1

log 1

PPU

!
2

R20 N0 EI0 
PCR

R0 e2 N0 PPU


m
PCR
1  log 1
T
R0 e2 N0

!
3

where N0 is the noise power spectral density, and PPU and

PCR are the transmit power of a PU and a CR, respectively.
In reality, PU activities may not be continuous and m is
not a constant, but a random variable. As the bursty trafc
is an interrupted Poisson process with onoff periods, we
assume PU activity follows a Poisson process with parameter k [19], the mean of the probability of inter-arrival time
is 1/k. In this sense, on average, CR nodes can be considered
to have v = T(1  x/k) time units without interfering with
PUs during [0, T], where x is the number of PU ows. Additionally, we can adjust the value of v to reect the spectrum sharing technique in (2) and (3). For example, if we
let m be T in (3), then (2) and (3) can represent a spectrum
sharing model where PUs and CRs can access the spectrum
resources at the same time and can avoid interference
without explicit signaling.
Based on (2) and (3), Fig. 3 shows two cases in which
adjusting the CR transmit power PCR or the radius of PER
can change the data rates for both CRs and PUs. These two
cases exemplify the impact of PER region on throughput. Besides, in order to get an optimal DPU we should choose a proper value of R/R0, which is 1.33 in the example shown in Fig. 3.
In Fig. 3(a), as PCR increases, DCR increases more quickly than
the declining DPU does. This is the reason that throughput always rises when PCR increases. In Fig. 3(b), DCR increases more
slowly than the declining DPU does. In fact, in order to choose
a proper value of PCR, we have to consider a feasible range of
PCR. A reasonable value of PCR can be chosen based on the
maximum transmit power for PPU and PCR regulated in the
current wireless network standards, such as Global System
for Mobile Communications (GSM) (where PPU is about
12 W), IEEE 802.22 (where PPU is less than 4 W [20]), IEEE
802.11 (where PCR is less than 100 mW), and IEEE 802.15.4
(where PCR is less than 100 mW).

Fig. 3. The normalized throughput of a PU and CRs versus PCR and R/R0,
when (a) m = 0.3 and (b) m = 0.7.

With the aforementioned discussion, we can see that

the PER region has signicant impact on the throughput
for both PUs and CRs. Furthermore, given a certain data
rate for a PU, C0, and a certain CR output power, PCR, we
can choose an optimal value of R/R0.

4. Proposed CM-MAC protocol

4.1. Overview of the proposed CM-MAC protocol
As discussed in Section 3, in order to meet the requirements of a CRAHN MAC protocol, we have to improve the
traditional CSMA/CA based MAC protocol with the frame
structure shown in Fig. 4(a). In Fig. 4(b), we use a CCC in order to exchange the control frames such as RTS frames, CTS
frames, and acknowledgement (ACK) frames. Following the
MAC protocol data unit (MPDU) transmission, a node will
wait for a short inter-frame space (SIFS) period and then
transmit the ACK frame. Before sending a RTS frame, the
spectrum sensing process will be initiated by a CR to make
sure there is a data transmission link on a certain channel k.

118

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

(a)

(b)

Fig. 4. Frame structures of (a) the traditional CSMA/CA-based MAC protocol and (b) the proposed CM-MAC protocol.

There are two advantages to using a CCC in CRAHNs.

First, possible collisions of control frames and data frames
can be avoided. Second, when spectrum availability
changes, assigning a CCC can alleviate the communication
efforts required to consult other CRs in a new spectrum
band for exchanging control information. The CCC discussed in the paper is used for exchanging the control
information, where the time spent on negotiating a channel for information exchange will be reduced.
The transmitter spectrum sensing (TSS) and receiver
spectrum sensing (RSS) procedures are employed, which
are dedicated to ensuring spectrum availability of links
for upcoming data transmissions. Checking the spectrum
availability before transmission can avoid transmission
failures. TSS is done by a CR transmitter and this CR transmitter combines the spectrum information into the immediate RTS frame eld. Meanwhile, RSS is completed on a
receiver and it combines the spectrum information into
the CTS frame. After the broadcasting stage of RTS/CTS
frames with piggybacked spectrum information, the neighboring CRs of the transmitter and receiver obtain the local
knowledge of one-hop spectrum availability.
It is worth noting that as we integrate the spectrum
information into the RTS/CTS routine, the update frequency of spectrum information on neighboring CRs is
dependent on the RTS/CTS request frequency (i.e., the data
transmission load). It is expected that in the saturated
mode of a CRAHN (i.e., a CR always has data payload to
send), the spectrum information can be frequently updated. For CRAHNs with less data load, the spectrum information may be updated less frequently, subject to the
possible failures of data transmissions caused by inaccurate spectrum information on the links.
Another solution to informing CRs of spectrum availability is to use a periodic updating mechanism that maintains broadcast frames containing spectrum availability

information. Because this solution may cause collisions

with the routine control frames and may result in signicant delays, this solution will not be considered in this
paper.
4.2. Channel aggregation technique
The separation of CCC and data channel cannot signicantly improve the throughput. This is because the data
transmission channel cannot be utilized at all before a successful RTS/CTS process.
A feasible way to further improve the throughput is to
decrease the transmission time of a data frame. We will
employ the channel aggregation in a similar manner to
the method suggested in [4,21]. An example of channel
aggregation is shown in Fig. 5(a), where, compared with
Fig. 4, the transmission time for a data payload in a MPDU
is reduced as the MPDU is split into three segments and
transmitted on three channels simultaneously. In each segment, we add the sequence number for each split data
payload.
It should be noted that the channels used in this technique are dependent on the spectrum sharing scheme. In
Fig. 5(b), we can see that the actual channels for transmission can be obtained after the negotiation stage. This negotiation stage can be either a RSS/TSS procedure or the
SPEC_CHANGE notication procedure in MSA algorithm,
which will be discussed later. Because the channel aggregation is employed, the sender is expected to receive three
ACKs for three split MPDUs.
4.3. Spectrum access and sharing
In the proposed CM-MAC protocol, a simplied spectrum access scheme is employed where a CR will access
the minimum available channels that can meet a certain

119

MPDU
#2-2

A
C
K

MPDU
#2-3

SIFS

A
C
K

A
C
K

Channel k

SIFS

MPDU
#2-1

A
C
K

Channel k+1

SIFS

MPDU
#1-3

SIFS

MPDU
#1-2

A
C
K

SIFS

MPDU
#1-1

SIFS

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

A
C
K

Channel k+2

(a)
i thCR

(i +1)th CR

Negotiation Stage:

TSS/RSS

Get agreement on the

transmission channels

SPEC_CHANGE

Transmission Stage:
Transmit data on the
channels

ACK(1)
ACK(2)
ACK(3)

(b)
Fig. 5. An example of channel aggregation in the view of (a) the MAC frame and (b) the sequence diagram.

rate, DCR, on a link between ith CR and (i + 1)th CR. Therefore, the set of channels accessed by a CR on this link is
PjKi;i1 tj
Ki,i+1(t) = Ki(t) \ Ki+1(t) and DCR k1
rk, where r(k)
is the rate supported on channel k. Besides, if we consider
the case where a CR uses all available channels to meet the
rate DCR, the link will have |Ki,i+1(t)| available channels for
data transmissions.
For spectrum sharing, instead of using the central coordination in IEEE 802.22 standard [22], we employ the distributed spectrum information exchange. A main goal in
the proposed CM-MAC is to ensure successful next onehop transmission. Therefore, it is necessary to show the
convergence of spectrum information exchange of the
TSS/RSS procedure.

For example, Fig. 6 illustrates that after the TSS procedure, CR nodes 2, 4, and 6 have the updated spectrum
information of CR node 1; and after RSS procedure, CR
nodes 1, 3, 5, 7 can receive updated spectrum information
from node 6. Although CR nodes 2 and 4 cannot get the updated spectrum information of CR node 6, it is not a problem as CR nodes 2 and 4 are not on the next transmission
link of node 6. The candidate CR nodes for the next transmission are CR nodes 1, 3, 5, and 7. As such, we can see that
the TSS/RSS procedure integrated in RTS/CTS/ACTS handshaking is sufcient and no signicant communication
overhead is required. All neighboring CRs can receive the
spectrum information which assures the successful next
one-hop transmission.

2
7

RTS

CTS

1
4

(a)

ACTS

7
4

4
5

(b)

(c)

Fig. 6. An example of intermediate results of the spectrum sharing procedure after (a) a RTS transmission, (b) a CTS transmission, and (c) an ACTS
transmission. The dotted lines are transmission ranges of CR node 1 and CR node 6.

120

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

The time spent on spectrum sensing is not negligible in

the proposed CM-MAC protocol because a spectrum sensing process usually can take as long as 20 ls [11], which
is similar to a typical short inter-frame space (SIFS) duration. Although it is more desirable to reduce the number
of times of spectrum sensing processes when PUs are inactive, the TSS/RSS procedure with RTS/CTS is exible as PU
activity may not be known in advance. Furthermore, we
discuss about the effect of sensing errors on the TSS/RSS
procedure. We assume there are no sensing errors during
the TSS/RSS procedure in the paper; however, the occurrence of sensing errors caused by a spectrum sensing function may exist during the transmissions of TSS/RSS related
frames. If there is a spectrum sensing error caused by the
spectrum sensing function, the data rate will be affected.
From [23], the sensing error may affect the data rate if a
transmission fails when a CR falsely identies an idle channel and the delay arisen by the re-detection efforts. However, the sensing errors can be mitigated by physicallayer (PHY) techniques [24,25], sensing scheduling protocols in an upper layer [26], and multiple access protocol
design [27]. For example, in the TSS/RSS procedure, if a
spectrum sensing error happens, it can be mitigated by
using the PHY techniques or rescheduling schemes of the
spectrum sensing during RTS/CTS/ACTS transmissions in
the link layer.

4.4. Mobility support algorithm

Because CRs in a CRAHN are able to move in the network and cause signicant interference to PU trafc, we
have to consider the case that CRs may move into a PER.
The negative effects when a CR moves into a PER include:
(1) PU communications experience interference from CRs
and (2) CR communications experience interference from
PU communications. Both effects result from the case that
a PU is not aware of the spectrum band being used by CRs.
As such, we need an algorithm to solve these issues.
One issue here is how a CR can readily know its vicinity
to a PER region. We propose to use the radio signal
strength indicator (RSSI) at the PHY layer to solve this
problem. The RSSI value is proportionally to the radio signal strength, with the beacons received from PU transmissions, a SU can know its vicinity to the PU. Although an
RSSI itself usually cannot give accurate distance estimations and it may cause some false alarms. However, we
considered the use of RSSI as a low-cost method to obtain
the proximity of a node to a PER and assume the sufcient
accuracy of the RSSI method, which can be improved by
triangulation or more sophisticated schemes. As the signal
is received by an ith CR on the CCC, RSSI(i, j), is inversely
proportional to the distance (d) between the ith CR and
the jth PU , the RSSI value can readily indicate the vicinity
to a PU if RSSI(i, j) is close to a constant threshold RSSIthres.
If we assume all the PUs have the same transmit power, the
RSSIthres value is sufcient for all CRs. If we assume that the
(i + 1)th CR node is communicating with the ith CR node,
we can describe the proposed MSA as follows:

Mobility Support Algorithm (MSA)

Input: RSSI(i, j), State(i), Ki,i+1(t), Kj(t)
for each CR
if RSSI(i, j) > RSSIthres AND State(i) == MAC_OPER
Kj(t)
func_SS(j)
if Kj(t) e Ki,i+1(t)
if State(i) == MAC_TRANSMIT
if |Ki,i+1(t)| ==1
send a STOP frame over CCC to the (i + 1)th node
on channel
the (i + 1)th node will stop the data
transmission
else if |Ki,i+1(t)| > 1
Ki,i+1(t)
{k|k e Ki,i+1(t), k R Kj(t)}
send the SPEC_CHANGE frame with Ki,i+1(t) to
the (i + 1)th node
end if
end if
if State(i) == MAC_IN_TRANSMIT
send a STOP frame over CCC to the (i + 1)th node
the (i + 1)th node will record the data frames/
segments already transmitted
the (i + 1)th node will reinitiate the
transmission for the remaining frames
end if
if State (i) == MAC_CTS/ACTS
Ki,i+1(t)
{k|k e Ki,i+1(t), k R Kj(t)}
send a CTS or ACTS frame piggybacking Ki,i+1(t)
to the transmitter over CCC
end if
State(i)
MAC_PER
end if
end if
if RSSI(i, j) 6 RSSIthres AND State(i) == MAC_PER
State(i)
MAC_OPER
end if
end for

State(i) records the current MAC state on a CR. If the ith

CR is in a PER region then State(i) = MAC_PER; if the CR is
out of a PER then State(i) = MAC_OPER; if the CR is in a
CTS/ACTS
procedure
with
a
transmitter,
then
State(i) = MAC_CTS/ACTS; if a CR is transmitting data
frames, then State(i) = MAC_TRANSMIT; State(i) = MAC_
IN_TRANSMIT means some frame segments have been
transmitted with the channel aggregation technique;
func_SS(j) is the spectrum sensing routine to sense which
channel is occupied by the jth PU. The STOP frame contains
short control information on the current channel, while the
SPEC_CHANGE frame contains the available channels for
transmission. When the (i + 1)th CR receives a SPEC_CHANGE frame, it then uses the available channels to send
the data frames. If a CR is in the process of sending CTS/
ACTS frames, then an updated CTS/ACTS will be resent to
the transmitter/receiver with the updated channel
information.

121

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

From the MSA algorithm, we can see that once the CR is

in the PER region, the data transmission should immediately stop or else retransmissions of frames may be caused.
When the state is MAC_IN_TANSMIT, meaning the frames
or frame segments have been in the process of transmission, the CR in a PER region should notify the transmitter
immediately in order to resume the transmission of
remaining frames or frame segments on the other CR.
5. Throughput analysis
This section provides the throughput analysis of the
proposed CM-MAC protocol.
In our analysis, as the PU topology can affect the performance in terms of throughput, we assume that the center
of each PU network will be at a distance of at least 2R. An
example of this CRAHN is shown in Fig. 7.
5.1. Average time regarding mobility
The mobility should be considered in the throughput
analysis as it affects the time spent on spectrum sensing
and MSA algorithm. We let the coverage of a CR be SCR,
where kSPER k > kSCR k, and we assume all the CRs have
identical coverage disk in the CRAHN. CR nodes are deployed in the disk area SPER (with radius R), following a
homogeneous Poisson process with density h per unit area.
When a CR node moving into the PER, it will run the
MSA algorithm. We are interested in the number of moving
nodes in an annulus area with radius [R0  r0, R0 + e + r0]
shown in Fig. 7. The number of neighbors of a node (i.e.,
the average degree) [28] between the CRs inside the PER
and the CRs outside the PER at a distance r0 can be obtained, i.e.,

EDeg 2ph

R0 er 0

PKi; i 1jsi; i 1sds

R0 r 0

where K(i, i + 1) means the event that ith CR and (i + 1)th

CR has a radio link while s(i, i + 1) is the distance between
them. From [28], if we assume s(i, i + 1) = r0

PKi; i 1jsi; i 1

1 1
10a
r0
 erf p log
bth
2 2
2#
10a10dB

!
5

where bth is the threshold value of the received power to

maintain the radio link; and # is the shadow fading variance; a is the path loss exponent.
When mobility of CRs happens, there is a probability of
retaining connectivity between CRs outside the PER, and
we know that probability P(Deg > 0) = 1  eE[Deg]. By algebraically manipulating (4) and (5), we can determine the
value of r0.
Next, we assume all CRs are moving and we consider
the case that at time t, a CR in the area with radius
[R0  r0, R0 + e + r0] just moved in the PER, the average time
spent regarding MSA is

ET MSA  P0 T SS P 11 T STOP P 12 T S

CHANGE P 1

T STOP P2 T CTS P3

where P 1 Pstate MAC TRANSMIT; P 2 Pstate MAC

IN TRANSMIT; P3 Pstate MAC CTS=ACTS, P11 is the
probability of sending STOP frame, P12 is the probability
of sensing SPEC_CHANGE frame. The average time spent
on one shot of spectrum sensing is T SS ; TSTOP and TS_CHAGE
are the time spent on transmitting the STOP and
SPEC_CHANGE frames, respectively. P0 is the probability
that the ith CR just moves in a PER region, we can obtain

P0 Pstatet MAC PERjstatet  1 MAC OPER:

Because it is difcult to give an exact value of P1, P2, or
P3 due to the fact that they are application-specic, we use
an estimated value for each of them.
For P1, we can take T TTdataCTS x as its value, where x is the
data

delay and empty slot time to consider.

T CTS
For P3, we can take T T
as its value. P2 is difcult
CTS x
data

to know because when there is a bulk of data to send, P2

is large; when there is a small bulk of data to send, P2 is
small. However, we can know the maximum value of P2,
i.e., T TxCTS x.
data

Furthermore, as P0 is usually dependent on the mobility

pattern of the CRs, if we assume the CRs are moving
following 1-dimensional (i.e., 1-D) correlated random
walks with bounds [0, 2R] with equal probability of
moving in two opposite directions, the steady-state probability at any location is 1/4R [29] if the speed is one-unit
length per time unit, which can be used to estimate the
value of P0.

5.2. Throughput analysis

Fig. 7. An example of a CRAHN with PUs and CRs.

The link throughput performance for a CR is considered

in this paper, and we use the normalized throughput dened in [30] as:

122

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

Fig. 8. Description of a successful data transmission.

EPayload transmitted in a slot time

Elength of a slot time

If a CR has a successful transmission as shown in Fig. 8,

it should be noted that the spectrum availability may
change because of the PU activity. As such, we let the time
spent during the TSS procedure and data transmission be
T ct , and the time spent during the RSS procedure and data
transmission be T cr .We can obtain:

T ct T CTSRTS SIFS D SIFS T RSS

T cr T CTS SIFS D

where D is the propagation delay.

Moreover, for the PU activity following a Poisson process (fNt; t P 0g with rate parameter k) during the time
interval [0, Ts]. Thus, we denote by Pre(k) the probability
that an available channel will be taken by a PU activity
on data channel k during Ts. If the number of PU ow at
a timeframe is larger than zero, the data channel k will
be taken by PUs and in this case we have

Pre k PNt T s  Nt > 0 1  ekT s

10

Apart from P re k, probabilities that affect the length of

a slot time include:
(a) The probability that no CR is transmitting, 1  Ptr,
where Ptr is the probability that there is at least
one transmission in the considered slot time.
(b) The probability that a data payload is successfully
transmitted, PtrPs, where Ps is the probability of a
transmission occurring on the channel is successful.
(c) The probability that a data payload is not successfully transmitted because of a collision Ptr(1  Ps).
Based on the aforementioned discussion, we can determine the throughput in (11) from the denition in (7).
From (7), the numerator can be easily deducted as it is
the length of payload; the denominator is determined by
the frame structure and aforementioned probabilities associated to the estimated length of each chunk of the frame,
as well as the possibility of transmitting a packet which is
determined by P0.

of a collision; r is the empty slot time; P0 is the probability

that CRs are moving into PER regions (P0 < 1); TMSA is the
average time length spent on the MSA algorithm when
mobility occurs and the data transmission time after running MSA algorithm; and P is the data frame length (i.e.,
the length of MPDU). Moreover, from Fig. 8, although we
have a CCC and the other channel for data transmission,
we can combine the factors on these two channels together
so that

T s DIFS T RTSCTSACTS 4D 4SIFS T RSS T TSS

T data T ACK

12

where TRTS+CTS+ACTS = TRTS + TCTS + TACTS, while TTSS and TRSS

are spectrum sensing time of transmitter and receiver,
respectively; Tc is related to the RTS frame collision on
CCC as

T c SIFS D DIFS

13

At this point, from the assumption that the data frames

have the same length (i.e., E[P] = P), and the expressions of
functions Ptr and Ps of p, where p is the stationary probability of a frame transmission by a CR, as well as Eqs. (8)(13),
we can get the average throughput result as

Pf


CT s
1  P0 r T c  rf0 T s  T c f 1C
P0 T MSA
14

where f np1  pn1 ; f0 1  pn ; C 1e

, and n is the
ekT s
number of CRs for transmission.
If we suppose the payload in MPDU will be transmitted
on the available channels on the link between the ith CR
and (i + 1) CR, the time spent on each available channel
can be |Ki,i+1(t)| times less at most (when each available
channel has the same bandwidth). However, in order to
assemble the split data frames on the receiver side, we
keep the same MPDU header for each available channel.
Therefore, the average time length on a data frame transmission is
kT s

T 0data T data u

1u
jKi;i1 tj

15

P s P tr EP
P re kT s
1  P 0 1  Ptr r P tr P s T s Ptr 1  P s T c 1P
P 0 T MSA
re k

11

where Ts is the time length of a successful data transmission, Tc is the time length that a channel is busy because

where u is the ratio of header length to payload length in

MPDU, which is usually less than 1.
The new throughput will be derived if we substitute
the Tdata with T 0data in (12). We can express g as
gn; p; k; j Ki;i1 tj; P0 .

123

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

Table 1
Parameter values for evaluation.
Parameter

Value

Parameter

Value

MAC data payload

MAC header
PHY header
RTS payload
CTS payload
ACTS payload
SIFS
DIFS
Slot time
Channel bit rate
ACK length

8184 bits
272 bits
128 bits
160bits + PHY header
112 bits + PHY header
112 bits + PHY header
20 ls
120 ls
50 ls
1 Mbps
112 bits + PHY header

Propagation delay (D)

No. of spectrum bands (K)
PHY max transmit power
PHY sensitivity
Rx spectrum sensing time (TRSS)
Tx spectrum sensing time (TTSS)
Empty slot time (r)
Receiving threshold power (bth)
Path loss exponent (a)
Dynamic operating range (r)
Stationary probability of a data transmission by a CR (p) (saturated mode)

1 ls
6
100 mW
100 dBm
20 ls
20 ls
50 ls
50 dB
4
1000
0.02

0.7

=0.0015

PU Poisson traffic (parameter )

Saturated mode for CR traffic

CM-MAC
CSMA/CA RTS/CTS MAC
SCA-MAC
=0.001 =0.002
=0.003 =0.004

0.6 (Kp=1; P0=0.5)

0.25

0.5

0.2

Throughput

Throughput

N=50
N=20
N=10

0.4

0.3

0.15

0.1
0.2

0.05
0.1

0
0

10

20

30

40

50

60

70

80

90

0.5

100

Number of CRs

(a)
0.18

( pc = 0.3; = 3; Kp=1; P0=0.5)

0.16

CM-MAC
CSMA/CA RTS/CTS MAC
SCA-MAC

0.35

=0.0015
0.3

0.1

Throughput

Throughput

=0.002, =0.0015

0.12

0.08
0.06

1.5
x 10-3

Fig. 10. CR link throughput versus N and k .

=0.002, =0.0001

0.14

PU Poisson traffic (parameter )

CR Poisson traffic (parameter )

N=50
N=10
N=20

0.25

0.2

0.15

0.04

0.1

0.02
0

20

40

60

80

100

0.05

Number of CRs

(b)
Fig. 9. Theoretical results of CR link throughput in the (a) saturated mode
and (b) non-saturated mode.

0.5

Fig. 11. CR link throughput versus N and k.

1.5
x 10-3

124

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

5.3. A special case of the proposed CM-MAC protocol

of transmission probability. Moreover, for the Poisson traf0

c model, q PfN 0 t 1g 1  ek T . The value of p0 can
be calculated by the collision probability pc and the total
number of stages q, as well as q.
We can obtain the retransmission probability on a
channel k, P 0re k, as follows

If CR trafc model follows a Poisson process

fN 0 t; t P 0g with parameter k0 , we can conduct the similar analysis in order to estimate the CR link throughput in
this case. Compared to the aforementioned analysis in the
saturated-mode case (i.e., a CR always has a data payload
to send), the special case we are discussing now is the
non-saturated mode (i.e., a CR does not always have a data
payload to send). The introduction of Poisson trafc model
will impose two changes to the aforementioned analytical
model. One is the stationary transmission probability p of a
data frame; the other is retransmission times of a data
frame.
We denote by p0 the new stationary transmission probability of a data frame. From [31], by assuming each CR has
a frame buffer and the probability of a frame arrival is q,
the non-saturated mode of CRs will nally affect the value

0.7

Pre k 1  1 

U CR  n
1
1  U CRPU r m 1

17

PU Poisson traffic: =0.002

CR Poisson traffic: '= 0.0001
Mobility probability: P0=0.5

0.18

K =0
P
K =1
P
K =2
P
K =3
P
K =4
P
K =5

0.5

16

As such, we can derive the estimated link throughput

results when we consider Poisson trafc model for both
PUs and CRs.
However, if there is no MSA algorithm and TSS/RSS procedure, the retransmission probability, P re k, can be calculated by the channel availability and spectrum hole
sufciency [4]. If we take the channel aggregation factor
as one, we can obtain [4]:

0.2

PU Poisson traffic: =0.002

Mobility probability: P0=0.5

0.6

0.16
0.14

Throughput

0.4

0.3

KP = 0
KP = 1
KP = 2
KP = 3
KP = 4
KP = 5

0.12
0.1
0.08
0.06

0.2

0.04

0.1
0.02

10

20

30

40

50

60

70

80

90

100

20

40

60

80

Number of CRs

Number of CRs

(b)

(a)

0.2
PU Poisson traffic: =0.002
CR Poisson traffic: '= 0.0015
Mobility probability: P0=0.5

0.18
0.16
0.14

Throughput

Throughput



P0re k P Nt > 0jN0 t > 0 1  ekTs

KP =0
KP =1
KP =2
KP =3
KP =4
KP =5

0.12
0.1
0.08
0.06
0.04
0.02
0

20

40

60

80

100

Number of CRs

(c)
Fig. 12. CR link throughput performance with different values of Kp in the (a) saturated mode, and (b)-(c) non-saturated mode.

100

125

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

where r is the dynamic operating range (i.e., the number of

channels a PU is operating on), UCR is the channel utilization of a CR, UCR+PU is the channel utilization by the PU
and CR, and m is the aggregation factor on the channel k.
Therefore, we can substitute the variable C in (14) with
(17). In this sense, we can compare the proposed CM-MAC
protocol with SCA-MAC protocol, because SCA-MAC belongs to the CSMA/CA RTS/CTS based MAC which has the
same nature of the proposed MAC.

be compared in the same condition. The essential parameters represented in Table 1, for example, are i.e.,
/ 0:03; P 8584 bits, TRTS+CTS+ACTS = 768 ls, Tc = 141 ls,
and TACK = 240 ls.
With the parameter values, we can get T s 1151:03
7938:48
jK
. Then, we can derive (14) as
i;i1 tj

g0

8584


@91f0
f

1
257:6P0 50 1151:037938:48KK



A 1010:03 7938:48
KK P

C=1C

6. Numerical results

18

where K is the total number of available spectrum bands, n

and C are dened in (14), and Kp = K  E[|Ki,i+1(t)|], i.e., the
average number of channels occupied by a PU. For the nonsaturated mode throughput, note that the variable C and p
will change as mentioned in Section 5. Moreover, it is expected that the larger the value of k, the more frequent
the PU trafc occupies the available spectrum bands. Note
that the throughput g is dened as the probability of

0.3

0.3

0.25

0.25

0.2

0.2

Throughput

Throughput

This section shows some numerical results based on the

aforementioned analysis. The parameters are listed in
Table 1. Besides, all the switching intervals from transmitting to receiving are set to zero. We assume that the number of CRs, N, is identical to n in (14), and these CRs are
transmitters in a CRAHN and can interfere with each other.
Moreover, the channel aggregation is not considered in
SCA-MAC and CM-MAC in order to let the three protocols

P0 =0.0

0.15

P0 =0.1
P0 =0.3

0.1

PU Poisson traffic: =0.0015

CR Poisson traffic:
0.0001

P0 =0.0

0.15

P0 = 0.1
P0 =0.3

0.1

P0 =0.5

P0 =0.5

P0 =0.7

P0 =0.7
0.05

0.05

PU Poisson traffic: =0.0015

0

20

40

60

80

100

20

40

Number of CRs

60

80

100

Number of CRs

(a)

(b)
PU Poisson traffic: =0.0015
CR Poisson traffic: '= 0.0015

0.3

Throughput

0.25

0.2
P0= 0.0
P0= 0.1

0.15

P0= 0.3

0.1

P0= 0.5
P0= 0.7

0.05

20

40

60

80

100

Number of CRs

(c)
Fig. 13. CR link throughput performance versus P0, where CR trafc is in the (a) saturated mode and (b)-(c) non-saturated mode with PU trafc.

126

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

successful transmitted frames per frame time. Furthermore, we will compare CM-MAC protocol with the aforementioned special case of CM-MAC protocol.
Fig. 9(a) shows the throughput performance versus the
number of CRs (assumed to be in the saturated mode)
when we take the value KP = 1 for all links. Using Kp = 1
for all nodes is to make the three protocols be compared
in the fair condition. For example, the CSMA/CA RTS/CTS
MAC is the baseline which can support the data communication in one spectrum band. The CR throughput decreases
when the value of k increases. This is because the PU trafc
with increasing k has a high possibility of affecting TSS/RSS
procedures. In this case, the PU becomes more active
which reduces CR access. Moreover, as shown in Fig. 9(a),
with any given value of k and N, the throughput performance of CM-MAC outperforms that of SCA-MAC. The reason for this result is that the delay caused by MSA
algorithm and TSS/RSS procedures in CM-MAC is smaller
than that of SCA-MAC. Furthermore, when the PU trafc
is heavy, CM-MAC can successfully reduce the effect of
the existence of PER regions. Besides, if we consider the

non-saturated mode for CR trafc (Fig. 9(b)), we can see

that the CM-MAC still outperforms CSMA/CA MAC and
SCA-MAC protocols.
Fig. 10 shows how the CR throughput changes versus N
and k0 . We can see that, when the CR trafc intensity increases, the throughput curves rise, reach a maximum,
and then decline. Moreover, in Fig. 10, when N increases
from 10 to 20, the CR throughput increases as well. However, in Fig. 10, when N = 50, the throughput sharply increases when k0 is small and decreases faster than the
throughput curves. The reason of this phenomenon is that
having more CRs will increase the trafc which increases
the chances of more data transmission conicts. This results in decreased CR throughput.
Fig. 11 shows how the CR throughput changes versus N
and PU Poisson trafc parameter, k. It can be seen that the
throughput curves decline with the increasing intensity of
PU trafc. Furthermore, in Fig. 11, when N increases from
10 to 20, the CR throughput increases correspondingly;
however, when N = 50, the overall throughput is slightly
less than the throughput when N = 20. This is expected

0.12

Response time (s)

0.1
0.08
0.06
0.04
0.02
0

10

20

30

40

50

60

Network size

(a)
CM-MAC (Simulation)
CM-MAC (Simulation w/ mobility)
CM-MAC
CSMA/CA RTS/CTS MAC (Simulation)
CSMA/CA RTS/CTS MAC (Simulation w/ mobility)
CSMA/CA RTS/CTS MAC

Throughput (success rate%)

60
50
40
30
20
10
0

10

20

30

40

50

60

Network size

(b)
Fig. 14. Simulation results. (a) Response time and (b) throughout performance.

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

because the increasing number of CR nodes results in

increasing conicts in the data transmission, which affects
the throughput performance.
Fig. 12 shows how the throughput performance can be
affected by the different values of Kp (i.e., the number of
exclusive channels occupied by PUs). Fig. 12(a) shows the
analytical results when CR Poisson trafc is in the saturated mode. The results for Poisson CR trafc (i.e., non-saturated mode) are presented in Fig. 12(b) and (c). We can
see that when we consider different intensities of CR Poisson trafc (reected by k0 ), the throughput performance
will change accordingly. Furthermore, we can see from
Fig. 12(a)(c) that when the number of available spectrum
bands to CRs increases (i.e., Kp decreases), the throughput
performance improves correspondingly. These results
meet our expectations because with additional channels
available to CRs, the overall throughput will increase.
Fig. 13 displays how the CR mobility factor, P0, affects
the CM-MAC throughput performance. Fig. 13(a) mainly
shows the MSA algorithm when the CR trafc is in the saturated mode; Fig. 13(b) and (c) show the throughput performance results for CR Poisson trafc. It is clear that P0
only slightly decreases the throughput performance because the MSA can deal with the mobility that affects the
disruption of the data transmissions. Therefore, we can
conclude the proposed CM-MAC is robust in the CR mobility case.
The simulation results are shown in Fig. 14, where a
CRAHN is randomly deployed in a square area and the
average speed of CR nodes is around 5 m/s. Fig. 14(a)
shows that our proposed protocol has a slightly longer
average response time than the CSMA/CA RTS/CTS MAC.
This is because we employed the ACTS and some spectrum
management features in the protocol. In Fig. 14(b), as the
essential features have been captured by the analysis, the
simulation throughput results have the same tendency as
the analytical results. Moreover, the mobility of the CR
nodes in CM-MAC has slighter effect than that of the
CSMA/CA RTS/CTS MAC. In the simulation, we used a speed
that can address the P0 in the analysis. However, we have
to make sure the nodes are moving within the deployed
area in order to model the case in the analytical discussion
where CR nodes can switch a spectrum band to maintain
the data transmissions. Furthermore, if a different channel
bit rate is used, the simulation results are expected in the
same tendency as the link throughput is not mainly dependent on the channel bit rate. Besides, we can expect the
performance of other CSMA/CA based protocols like the
IEEE 802.15.4 MAC protocol, which may have a better response time when RTS/CTS mechanism is reduced.

7. Conclusions
In this paper, we introduced the CM-MAC protocol, a
MAC protocol for CRAHNs, by mainly considering CR
mobility of CRs and PUs PER regions. We included the
spectrum sensing in the handshaking procedure, and thus
the spectrum information updates on CRs are highly
dependent on the PU trafc and the CR data trafc. Moreover, we demonstrated the effectiveness of CM-MAC by

127

showing the analytical link throughput, which is mainly

related to the following parameters: number of CRs, stationary probability of a frame transmission by a CR, probability of CR mobility, PU and CR trafc, and the set of
available channels. The results showed that the CM-MAC
protocol outperforms the IEEE 802.11 MAC and SCA-MAC
protocols in terms of throughput performance. The results
have shown that the proposed algorithm is robust with respect to CR movements. This result makes CM-MAC suitable for CRAHNs. Although only a 1-D random walk was
discussed as a movement model for CRs, the proposed
CM-MAC can be applied to a more general case like CRs
with 2-D movements. We will investigate 2-D movements
in future works, and the results in more realistic channel
conditions will be investigated.
References
[1] I.F. Akyildiz, W.-Y. Lee, K.R. Chowdhury, CRAHNs: cognitive radio ad
hoc networks, Ad Hoc Netw. 7 (2009) 810836.
[2] I.F. Akyildiz, W.Y. Lee, M.C. Vuran, S. Mohanty, A survey on spectrum
management in cognitive radio networks, IEEE Commun. Mag. 46
(2008) 4048.
[3] V. Mai, N. Devroye, V. Tarokh, On the primary exclusive region of
cognitive networks, IEEE Trans. Wireless Commun. 8 (2009) 3380
3385.
[4] A. Chia-Chun Hsu, D.S.L. Weit, C.C.J. Kuo. A cognitive MAC protocol
using statistical channel allocation for wireless ad-hoc networks, in:
WCNC07, 2007, pp. 105110.
[5] H.B. Salameh, M. Krunz, O. Younis. Distance- and trafc-aware
channel assignment in cognitive radio networks, in: SECON08, 2008,
pp. 1018.
[6] H.B. Salameh, M. Krunz, O. Younis, MAC protocol for opportunistic
cognitive radio networks with soft guarantees, IEEE Trans. Mob.
Comput. 8 (2009) 13391352.
[7] S. Tao, C. Shuguang, M. Krunz. Medium access control for multichannel parallel transmission in cognitive radio networks, in:
GLOBECOM06, 2006, pp. 15.
[8] N. Jain, S.R. Das, A. Nasipuri. A multichannel CSMA MAC protocol
with receiver-based channel selection for multihop wireless
networks, in: ICCCN01, 2001, pp. 432439.
[9] B. Sadeghi, V. Kanodia, A. Sabharwal, E. Knightly, OAR: an
opportunistic auto-rate media access protocol for ad hoc networks,
Wireless Netw. 11 (2005) 3953.
[10] Z. Hongqiang, W. Jianfeng, F. Yuguang, DUCHA: a new dual-channel
MAC protocol for multihop ad hoc networks, IEEE Trans. Wireless
Commun. 5 (2006) 32243233.
[11] C. Feng, Z. Hongqiang, F. Yuguang, An opportunistic multiradio MAC
protocol in multirate wireless ad hoc networks, IEEE Trans. Wireless
Commun. 8 (2009) 26422651.
[12] D. Cabric, R.W. Brodersen, Physical layer design issues unique to
cognitive radio systems, in: PIMRC05, vol. 752, 2005, pp. 759763.
[13] K.R. Chowdhury, M. Di Felice, I.F. Akyildiz, IEEE. TP-CRAHN: a
transport protocol for cognitive radio ad-hoc networks, in:
Infocom09, 2009, pp. 24822490.
[14] J. So, N.H. Vaidya. Multi-channel mac for ad hoc networks: handling
multi-channel hidden terminals using a single transceiver, in:
MobiHoc04, Tokyo, Japan: ACM, 2004, pp. 222233.
[15] C. Cordeiro, K. Challapali. C-MAC: a cognitive MAC protocol for
multi-channel wireless networks, in: DySPAN07, 2007, pp. 147
157.
[16] P. Pawelczak, R. Venkatesha Prasad, L. Xia, I.G.M.M. Niemegeers.
Cognitive radio emergency networks - requirements and design, in:
DySPAN05, 2005, pp. 601606.
[17] V.S. Frost, B. Melamed, Trafc modeling for telecommunications
networks, Commun. Mag. IEEE 32 (1994) 7081.
[18] J.M. Peha, Sharing spectrum through spectrum policy reform and
cognitive radio, Proc. IEEE 97 (2009) 708719.
[19] P. Tran-Gia, D. Staehle, K. Leibnitz, Source trafc modeling of
wireless applications, AEU Int. J. Electron. Commun. 55 (2001)
2736.
[20] C. Corderio, K. Challapali, D. Birru, S. Shankar, IEEE 802.22: an
introduction to the rst wireless standard based on cognitive radios,
J. Commun. 1 (2006) 3847.

128

P. Hu, M. Ibnkahla / Ad Hoc Networks 17 (2014) 114128

[21] D. Skordoulis, N. Qiang, C. Hsiao-Hwa, A.P. Stephens, L. Changwen, A.

Jamalipour, IEEE 802.11n MAC frame aggregation mechanisms for
next-generation high-throughput WLANs, IEEE Wirel. Commun. 15
(2008) 4047.
[22] C. Stevenson, G. Chouinard, L. Zhongding, H. Wendong, S.
Shellhammer, W. Caldwell, IEEE 802.22: the rst cognitive radio
wireless regional area network standard, IEEE Commun. Mag. 47
(2009) 130138.
[23] T. Shu, M. Krunz. Throughput-efcient sequential channel sensing
and probing in cognitive radio networks under sensing errors, in:
Proceedings of the 15th annual international conference on Mobile
computing and networking, Beijing, China: ACM, 2009, pp. 3748.
[24] D. Cabric, S.M. Mishra, R.W. Brodersen, Implementation issues in
spectrum sensing for cognitive radios, in: Signals, Systems and
Computers, 2004 Conference Record of the Thirty-Eighth Asilomar
Conference, vol. 771, 2004, pp. 772776.
[25] A. Ghasemi, E.S. Sousa, Collaborative spectrum sensing for
opportunistic access in fading environments, in: DySPAN05, 2005,
pp. 131136.
[26] I.F. Akyildiz, B.F. Lo, R. Balakrishnan, Cooperative spectrum sensing
in cognitive radio networks: a survey, Phys. Commun. 4 (2011) 40
62.
[27] A.A. El-Sherif, K.J.R. Liu, Joint design of spectrum sensing and
channel access in cognitive radio networks, Wireless Commun.
IEEE Trans. 10 (2011) 17431753.
[28] C. Bettstetter, C. Hartmann, Connectivity of wireless multihop
networks in a shadow fading environment, Wireless Netw. 11
(2005) 571579.
[29] S. Bandyopadhyay, E.J. Coyle, T. Falck, Stochastic properties of
mobility models in mobile ad hoc networks, IEEE Trans. Mob.
Comput. 6 (2007) 12181229.
[30] G. Bianchi, Performance analysis of the IEEE 802.11 distributed
coordination function, IEEE J. Sel. Areas Commun. 18 (2000) 535
547.
[31] K. Duffy, D. Malone, D.J. Leith, Modeling the 802.11 distributed
coordination function in non-saturated conditions, IEEE Commun.
Lett. 9 (2005) 715717.

Peng Hu received his M.S. degree in communication and information systems from
Wuhan University of Technology, Wuhan,
China and his Ph.D. degree from Queens
University, Kingston, Canada. He has served as
a reviewer for several international journals
and conferences including IEEE Transactions
on Mobile Computing, EURASIP Journal on
Wireless Communications and Networking,
and Wireless Communications and Mobile
Computing. His current research interest is in
the eld of cognitive radio ad hoc networks,
sensor networks, and Internet of Things.

Mohamad Ibnkahla received the Ph.D.

degree and the Habilitation a Diriger des
Recherches degree (HDR) from the National
Polytechnic Institute of Toulouse (INPT),
Toulouse, France, in 1996 and 1998, respectively. He joined Queens University, Canada,
in 2000, where he is now a full professor. He
led several projects applying wireless sensor
networks in various areas such as environment monitoring, wildlife tracking, pollution
detection and control, food traceability and
safety risk monitoring, highway safety, intelligent transportation, and water management. He has published Signal
Processing for Mobile Communications Handbook, CRC Press, 2004;
Adaptive Signal Processing in Wireless Communications, CRC Press, 2008;
Adaptation and Cross-layer Design in Wireless Networks, CRC Press, 2008,
Wireless Sensor Networks: A Cognitive Perspective, CRC Press, 2012, and
Cooperative Cognitive Radio: The Complete Spectrum Cycle, CRC Press,
2014.