Addressing the Multichannel Selection, Scheduling and Coordination Cost

International Journal on AdHoc Networking Systems ( IJANS) Vol. 5, No. 2, April 2015
DOI : 10.5121/ijans.2015.5201 1
ADDRESSING THE MULTICHANNEL
SELECTION, SCHEDULING AND
COORDINATION COST
Mthulisi Velempini1
and Mqhele E. Dlodlo2
1
University of Limpopo , South Africa
2
University of Cape Town , South Africa
Abstract
We discuss a new multichannel coordination and scheduling challenge called the Multichannel Scheduling
Cost (MSC). It is caused by the inability of the control channel to schedule simultaneous transmissions
resulting in data channels lying idle and their bandwidth underutilized. As a result, wasted bandwidth
increases as the number of data channels increases. To address this challenge, a cyclic scheduling
Algorithm (CSA) is proposed. It employs one dedicated signaling channel and 4 data channels. It is
premised on a proactive channel reservation scheme which reduces the idleness of channels. The scheme
ensures that data channels are reserved while they are still busy. The data channels are reserved while
their remaining transmission duration is equal to the virtual carrier sensing duration. This gives the next
pair sufficient time to reserve the current data channel before it becomes available, limiting the MSC to the
first cycle. The simulation results show a performance benefit of the CSA scheme in addressing the effects
of the MSC.
Keywords: Bandwidth Utilization; Connectivity; Control channel; Multichannel Scheduling Cost
1.0 Introduction
The implementation of multichannel MAC protocols in Wireless Mesh Networks (WMN) has
generated some encouraging results, which increases network capacity and throughput. However,
there are still outstanding challenges which degrade the capacity of multichannel networks. For
example, the channel assignment, selection and coordination strategies result in the creation of a
new multichannel degradation challenge we call the Multichannel Scheduling Cost (MSC). MSC
wastes the capacity of data channels during the reservation phase. Data channels remain idle for
long durations waiting for their turns to be reserved, resulting in the underutilization of their
capacity. The amount of wasted bandwidth increases linearly as the number of data channels
increases. Furthermore, the MSC challenge is recurring and repetitive. To address this challenge,
we propose a proactive data channel reservation scheme which addresses the MSC and improves
the utilization of data channels. The scheme reduces the idle durations of data channels while
improving their capacity.
Data channels are reserved in phases. The first channel to be reserved in the first cycle suffers the
least degradation caused by the MSC, while the last data channel suffers the worst degradation.
The adverse effect of the MSC increases as the number of data channels increases. The MSC is a
scheduling challenge is addressed in this paper through the implementation of a proactive
scheduling scheme which limits the MSC to the first cycle, thereby improving the performance of
multichannel MAC protocols.

2
The MSC interference challenge affects the multichannel MAC protocols which implement
control channel coordination schemes. Besides the perceived performance benefits of the control
channel schemes, they give rise to the MSC. We characterize and demonstrate the impact of the
MSC on network performance. A proposed MSC solution is presented and evaluated through
network simulations. The results show an improvement in the performance of the network.
The scheme is implemented in a backhaul of static mesh routers equipped with intelligence. The
word intelligence in this context means the ability of the network to acquire knowledge of the
network status, to facilitate distributed decision-making, to aid data channel reservation. The
number of channels was not changed in the simulation environment.
The paper is structured as follows: The next Section justifies the implementation of a common
signaling channel. Section 3 discusses related work. The system model is discussed and analyzed
in Section 4 and 5 respectively. The simulation environment is presented in Section 6 and the
results are analyzed in Section 7. Section 8 summarizes the paper.
2.0 Motivation
Multichannel MAC schemes implementing a common control channel have been designed. The
protocols show a significant improvement in network performance and they offer reasonable
capacity. However, their implementations have given rise to a new multichannel scheduling cost.
The MSC has not been addressed in literature. It is a challenge which degrades network
performance.
The use of multi-channels avail more bandwidth which is a requirement of the next generation
networks however, channel selection, access and coordination remains a challenge. Channel
coordination challenges result in the degradation of bandwidth. This challenge is evident in MSC
in which a number of channels remain idle while they wait for the active channel to finish its
transmission. Furthermore, channels are accessed in turns, one at a time successively leading to
repetitive cycles of bandwidth wastage. A scheduling algorithm which reduces either the idle
duration of channels or which addresses these idle slots is required. In this paper, an algorithm
which restricts the idleness of channel to the first cycle and eliminates them in the successive
cycles is presented
3.0 Related Work
Multichannel MAC protocols which employ a single transceiver and a control channel, which is
reserved for signaling, are examined. The impact of the MSC on network performance is
highlighted. The emphasis is on the MSC challenge and how it degrades the capacity of
multichannel MAC protocols.
The scheme in [1] employs a temporary control channel called the quiescent channel. The
protocol makes use of a dedicated busy tone channel to coordinate data channels access to
prevent data collision. The busy tone channel wastes bandwidth, though it does not give rise to
the MSC challenge it requires optimization.
The second algorithm proposed in [1] which divides a communication session into control and
data windows is subjected to the effects of the MSC. The MSC is experienced during the control
window. Given that the control window precedes any communication, the effects of the MSC are
recurring. Lastly, during the data window there is no manifestation of the MSC. All the channels
are utilized during the data window.

3
The protocol utilizes only one channel during the control window sessions. The data channels lie
idle and their bandwidth is wasted. It allows more than one pair of nodes to reserve one data
channel thereby worsening the effects of the MSC when nodes take turns to transmit on the
reserved data channels.
In [2], a window based scheme is proposed. Data channels are reserved during ATIM window
through the default channel. One data channel can be reserved by many nodes, which later
contend to access it during the data window. The scheme is similar to the one described above
however, the implementation of the virtual carrier sensing (VCS) results in the worst case of the
MSC which is repetitive both during the ATIM and the data windows. The two forms of MSC are
experienced degrading the network performance significantly.
A window based scheme which implements a temporary control channel is also proposed in [3].
The protocol consists of a contention reservation interval (CRI) and the contention free interval
(CFI). Terminals contend for data channels during the CRI and then defer their transmissions
until the start of the CFI. The deferment of transmission and the reservation of channels during
the CRI give rise to the MSC challenge.
The protocol proposed in [4] implements the concept of home channels instead of a control
channel. The implementation of the home channels addresses the MSC at the cost of partitioning
the network and lack of reach.
In [5], a single transceiver Multichannel MAC protocol which implements contention and data
windows is proposed. A three way channel reservation handshake is implemented with an
additional control packet called the Multichannel CTS Recognition (MCTS-R). The increasing
control packets worsen the effects of the MSC. Furthermore, the data channels are not used
during the control window and they are reserved in phases sequentially resulting in the
manifestation of the MSC challenge.
The Group Allocation Multi hop Multiple Access (GAMMA) proposed in [6] implements
multiple transmission channels. Each transmitter has a unique channel which reduces the effects
of the MSC. The process of realigning data slots, the management of group membership,
registration and de-registration procedures however, worsens the MSC.
The protocol proposed in [7] introduces the idea of distributed information sharing to ensure that
communicating pairs do not make independent decisions. The proposed protocol does implement
a dedicated control channel however; it is optimized to solve the terminal deafness problem and
collisions in the data channels. The protocol introduces up to six control packets which worsens
the impact of the MSC which is repetitive.
In [8] a Distributed Queue Dual Channel (DQDC) scheme is proposed. The DQDC scheme
introduces 4 control packets which increases the idle durations of the data channels which in turn
worsens the impact of the MSC challenge.
In [9] a scheme which implements a control channel and N traffic channels is presented. A node
first senses all the data channels including the control channel before it sends an RTS packet.
Both the transmitter and the receiver sense all the channels which increases the MSC. The effects
of the MSC have been worsened by an increase in the size of control packets.
In [10], an asynchronous Multichannel Coordination Protocol (AMCP) is proposed. The protocol
employs a single transceiver and a dedicated common control channel. Bandwidth is wasted when
the preferred channel of the sender is rejected by the receiver. The sender has to select randomly
the receiver’s suggested data channels and restarts the contention cycle. Furthermore, control

4
packets have increased in size. In addition, a new control packet, the Reserve (RES) has been
introduced. Bandwidth is further degraded as nodes attempt to reserve data channels more than
once. All these configurations contribute to the worst case of the MSC challenge.
In [11], a busy tone scheme is proposed. The scheme assumes that nodes can transmit and receive
simultaneously. A channel is divided into two sub channels, a data and a control channel. Busy
tone signals are transmitted on the control channel. The data channels are not used during data
channel reservation, which results in the MSC challenge.
In general, multichannel MAC protocols outperform single channel MAC protocols as
demonstrated in [12], [13]. The multichannel schemes have been investigated in cognitive
networks with a measure of success [14], [15]. They have also been implemented in vehicular ad
hoc networks [16], in wireless sensor networks [17], [18], [19], [20] and in general ad hoc
networks [21], [22], [23], [24], and [25]. This shows that multichannel technique is a promising
approach requiring further investigation.
4.0 System Model
In this section, we discuss the architectural design ideal for the implementation of the MSC
solution. The architecture is memory based which is equipped with nodes with sufficient network
status to facilitate the reservation of data channels.
The architecture is equipped with a network of mesh routers deployed in the ad hoc network of
mesh clients. The mesh routers are networked to facilitate communication within the network of
mesh clients. The architecture is configured such that each mesh router is within the transmission
range of the next mesh router. The mesh routers store network status information which is shared
with mesh clients to support their medium access and scheduling decisions. The proposed
architecture is depicted in Fig 1. The architecture consists of routers enclosed in blue lines. All
the Mobile nodes are assumed to be within the coverage of the network of mesh routers. The
network of support mesh routers provides a support layer to the mesh architecture.
The network support mesh routers maintain a data structure we call a Network Status Table
(NST) which is updated using overheard control packets. The network support nodes are referred
to as the Network Status Table Nodes (NSTN). The NSTNs store information on the availability
of data channels, the list of in use data channels and the duration of their unavailable – the inter-
phase time. A data channel is available when its remaining transmission time is equal to the VCS
duration of the next pair. The remaining time is determined by a hold off duration - the inter-
phase time technique which allows data channels to be reserved in a proactive manner.

5
Fig 1. The architecture of the proposed network support infrastructure in a hybrid mesh network
A new inter-frame space known as the inter-phase replaces the DIFS ensures cyclical
transmission of data. It is a proactive scheme which facilitates the reservation of data channels
before they finish their current transmission, effectively reducing the idleness of data channels
between the phases. The reduction of the idleness of data channels reduces the effects of the
MSC. When data channels finish their transmission in the previous phase while waiting for their
turn in the next phase to be reserved, they lie idle for long durations impacting negatively on the
efficiency of the protocol which results in MSC. Given our proactive scheme, the wait period is
reduced significantly which addresses the effects of the MSC.
A data channel is reserved when the inter-phase time has expired and nodes use the CSMA/CA to
contend for the control channel. The inter-phase replaces the DIFS implemented in IEEE 802.11
MAC and its value is stored in the NAV of nodes. It also doubles as the transmission duration of
data packets. The inter-phase time is a summation of data transmission duration plus one
switching time minus the control channel handshake [26] [27]. The inter-phase time is stored in
the NSTN as data transmission duration of the given data channel. The NSTNs maintain a
sequence of the data channels to ensure that they are reserved in a round robin basis. Furthermore,
the network status stored in tables is availed to mesh client node with insufficient knowledge of
the status of the network. The said node probes the nearest NSTN. The maintenance of the
network status tables is assigned to powerful nodes, the mesh routers. NSTN overhear control
packets, use overhead data to update their data structures. The network support nodes are
designed to equip mesh clients that cannot update their VANs after missing control packets.
Fig 2 demonstrates the functionality of the support scheme. The flow chart shows that depending
on the availability of the network, a given node can contend for the control channel, otherwise it
has to first probe the nearest NSTN for detailed network status. Three scenarios exist in which
nodes may not have adequate knowledge of the network. These are the joining, the returning and
start-up nodes.
A mesh client with insufficient network information probes the nearest NSTN before contending
for the control channel to reserves a channel. Network related data is sent to the probing node by

6
an NSTN. Overhearing nodes also update their tables; only if their tables are out dated. A node
which listens on the control channel is expected to be up to date; hence there is no need for it to
probe the NSTNs or to rely on probe responses.
Fig 2. The implementation of the network infrastructure in the Cyclic Scheduling Algorithm
5.0 Model Analysis
The MSC wastes bandwidth on all the data channels and it is repetitive. However, the proposed
scheme limits the MSC to the first cycle. This is achieved through the implementation of inter-
phase time scheme. The data channels are reserved when they are still busy transmitting current
data frames. The implementation of the inter-phase time ensures that data channels reservation is
done proactively when data transmission is about to end. Secondly, the Extended Inter-Frame
Space (EIFS) has been reconfigured and changed into a shorter Inter-Frame Space (IFS). In
multichannel MAC protocols, there is no need for nodes which receive erroneous packets to defer
for the EIFS duration. The EIFS worsens the MSC. The EIFS was designed to protect the ACK
packets in single channel MAC protocols. In multichannel MAC protocols, ACK packets are
transmitted on a different channel.
Fig 3 shows how the window based multichannel MAC protocols worsen the MSC. It also shows
the extent at which data channels are underutilized during the control window.
Start
Joining or Returning node or
new network
Node has knowledge of the
network
Contend for control channel
Select data channel
Switch onto data channel
Switch back to the
control channel
Transmit data packets
Next Node
Probe NSTN for network
status
Network available
Yes
No
Stop
No
Yes

7
Fig 3. The poor utilization of data channels in Window based multichannel MAC protocols.
The column is marked either as a control window or data window. The rows represent the
channels and the top row represents the control channel. During the control window, the control
channel is marked “Reservation” while in the data window it is marked “Data Frame”. Data
channels are reserved through the control channel during the control window and then as a data
channel inside the data window to transmit data frames. The last two rows depict the two data
channels. They are both marked “DATA IDLE” during the control window, demonstrating the
underutilization of their capacity during the control window. It shows the extent to which the
MSC affects the performance of the network.
Fig 4 characterizes and models the MSC. It also shows how the MSC challenge degrades the
capacity of data channels. It shows the severity of the MSC and the extent at which control and
data channels are underutilized. The vertical axis denotes the number of channels marked channel
0 to 7. The horizontal axis denotes the elapsed transmission time. The control channel is assumed
to be Channel 0 while the rest are data channels. The capacity of all data channels is underutilized
between T0 and T1 while nodes are busy contending for the control channel and reserving
Channel 1. Channel 2 is reserved between epoch T1 and T2, while all other channels lie idle
except channel 1 which is currently transmitting data. In essence, Channel 1 is underutilized
between T0 and T1, Channel 2, is underutilized between T0 and T2, Channel 3 is underutilized
between T0 and T3, channel 4 between T0 and T4, channel 5 is underutilized between T0 and T5,
while channel 6 is underutilized between T0 and T6. Lastly, channel 7 underutilized between T0
and T7.
As illustrated in Figure 4, the last channel to be reserved, channel 7 in this case is subjected to the
worst effects of the MSC while the first channel (which is reserved first in phase) suffers the least
degradation due to the effects of the MSC. On the other hand the control channel lies idle
between T7 and T8. Thereafter the pattern is repeated for Data channels from T8 to T14. This
shows that the effects of the MSC are repetitive and recurring. The MSC degrades severely the
performance of the multichannel MAC protocols as illustrated in Fig 4. The illustration clearly
demonstrates the significance of this work and the magnitude of the MSC challenge. The
proposed model restricts the MSC to the first phase, there after its effects are eliminated. It is not
possible to eliminate the MSC in the first phase, due to the fact that nodes cannot be reserved in a
proactive manner in the first phase. Secondly, data transmissions cannot be scheduled to all the
available data channels concurrently. Figure 5 characterize the MSC challenge.
Control Channel
Data Channel 1
Data Channel 2
Reservation
DATA IDLE
Data Frame
DATA IDLE
Data Frame
Data Frame
DATA IDLE
Data Frame
Reservation
Data Frame
Data Frame
DATA IDLE
Control Window Control Window
Data Window Data Window

8
Fig 4. The Multichannel Scheduling Cost
Fig 5 depicts the behaviour of the MSC as more data channels are added. It also shows how the
effects of the MSC increase with the increase in the number of channels. The 2nd data channel
incurs twice the degradation of the 1st data channel, while the degradation of the 3rd data channel
is thrice that of the 1st data channel. The degradation of the fourth data channel is 4 times that of
the 1st data channel. Thus, the effects of MSC increase with the increase in the number of data
channels. The total degradation of bandwidth is the summation of wasted bandwidth on all data
channels. This is a high cost which degrades severely the performance of multi-channel MAC
protocols. This cost requires an effective and efficient solution.
To calculate the wasted bandwidth due to the effects of the MSC, the following equations can be
employed:
MSC = wb + 2 x wb + 3 x wb + ....... + (n - 1) x wb + n x wb (1)
(2)
Where n is the number of data channels and wb is the amount of wasted bandwidth.

9
Fig 5. The Behaviour of the MSC as the channels increase in number
6.0 Simulation Model
The proposed scheme was implemented in NS 2. The performance of the scheme was evaluated
in a multi-channel frame work developed in [10]. The EIFS was reconfigured and changed to
shorter IFS called the CIFS. The CIFS is the summation of the transmission duration of the CTS
packet and the SIFS duration. Nodes defer for CIFS instead of EIFS upon receiving erroneous
control packets. In the proposed scheme, nodes defer for SIFS after the data channel has been
reserved. The two channel switching times have been added to data packet transmission
durations.
A proactive reservation strategy which reduces the MSC by improving the utilization of the
channels is implemented. The CIFS, the data channel reservations strategy and the deferring rules
are designed to reduce idle durations of channels. They improve the utilization of channels which
improves the performance of the proposed protocol. The scalability of the scheme and its
robustness in large networks with high incidents of interference was also evaluated.
A network with 5 channels, a control channel and 4 data channels was implemented. The
channels were assumed to be orthogonal and of equal bandwidth, which was set to 2Mbs. The
number of channels was changed however; the size of the network was varied between 4 and 30
nodes to evaluate the scalability of the proposed scheme. A Non Ad Hoc (NOAH) routing
algorithm was also implemented.
In the simulation, 5 different network sizes in terms of number of nodes were considered. The
network sizes with the following number of nodes were considered: 4, 6, 8, 10, and 30 nodes.
Simulations were run for each network size, and the topologies were assumed to be general. Data
packets were assumed to be 1000 bytes long and of type CBR. A single hop network set up was
assumed with packets sent directly to the receiver. The RTS and CTS packets were transmitted on
the control channel and the DATA and ACK packets were transmitted on the data channels. We

10
also took into account 2 channel switching times, each assumed to be 224µs. For detailed
simulation environment see the table below.
Table1. MAC Layer and other Parameter Settings
Parameter Value
SIFS 10 µs
DIFS 50 µs
EIFS 364µs
CIFS 56 (µs)
Slot time 20 µs
Data rate 2 Mbps
Basic rate 2 Mbps
Control Channel 1
Data Channels 4
Switching time 224µs
Data Type CBR
Packet size 1000 bytes
CW max 1023
CW min 7
Short Retry Limit 7
Long Retry Limit 4
PLCP Length 192 bits
PLCP rate 1 Mbps
RTS 20 bytes
CTS 14 bytes
ACK 14 bytes
MAC Header – Data 28 bytes
MAC Header rate Basic rate
Routing Agent NOAH
Network Sizes 4,6,8,10 and 30 nodes
PLCP Header 6 Bytes
Control Channel
Propagation Delay
2µs
Data Channel Propagation
Delay
4µs
PLCP Preamble 18 Bytes
7.0 Simulation Results
7.1 Analysis of dropped packets
In this segment, we analyze the number of dropped packets in evaluating the performance of the
proposed scheme. The emphasis is on the number of dropped packets caused by longer system
response times. The efficiency of service points is therefore investigated. In this case the service
time of the control channel is examined. The services time affects the number of dropped packets

11
from the buffer. This metric relates to the utilization of channels in the sense that, if the utilization
of the channel is poor, longer system response times are expected, and more packets are dropped
as a result. Improving the utilization of channels causes fewer packets to be dropped. The
improvement in the utilization of the channel and the dropping of fewer packets is a sufficient
indicator of the effectiveness of the proposed scheme in addressing the MSC. In Fig 7, the
scheme is evaluated in a network with 4 nodes.
Fig 7. Analyzing dropped packets in a network with 4 nodes.
The CSA dropped the least number of packets demonstrating that the scheme did experience
fewer and shorter delays as a result of improved channel utilization. The packets were not
subjected to longer waiting durations before they were processed. In CSA, the next pair does not
wait for the current pair to finish its transmission; it proactively reserves it during the course of
the ongoing data transmission. The improved service rates in the CSA effectively address the
effects of the MSC.
The nodes were increased to 6 in Fig 8. The nodes were grouped into three transmitters and three
receivers, generating three data flows. The CSA scheme did drop fewer packets as compared to
the AMCP. All the data flows dropped fewer packets, a clear indication that the proposed CSA
scheme is effective in addressing the MSC.
Fig 8. The performance of the CSA in a network with 6 nodes

12
In Fig 9, the network size was scaled up. The number of nodes was increased to 10. The number
of data flows was more that the number of the data channels. As a result, the contending fifth data
flow had an effect on the overall performance of the network. The two schemes did experience
the highest number of dropped packets, which degraded the performance of the two schemes.
Fig 9. The performance of the CSA in a general network with 10 nodes
In the nutshell, the proposed scheme performed better than the AMCP. However, it was
outperformed in the 1st data flow. This is due to its longer transmission durations and the effects
of the interfering fifth data flow. The CSA is scalable and its performance improves as the
network increase in size. This shows that the scheme addresses effectively the MSC challenge.
The results in Fig 10 validate this argument.
Fig 10. The total number of dropped packets in a network with 30 nodes
In Fig 10, the number of nodes was increased to 30 nodes. The ratio of data channels to data
flows was almost one to four (1:4). In any given time, there were about 4 data flows contending

13
for 1 data channel. The amount of interference was high and the network resources were limited.
The performance of CSA was superior and it improved as the size of the network was increased.
It dropped fewer packets in each of the transmitting nodes. The number of dropped packets was
less than the ones which were dropped in Fig 9. The results illustrate that the CSA scheme is
more scalable and effective in addressing the MSC.
7.2 Packet Drop Rate
In these last set of results, the frequency at which the packets were dropped is analyzed. The
reduction of the end-to-end delay shows that the proposed scheme is effective in addressing the
MSC. It shows to what extent the system service rate and response time affect the behavior of the
queue and the buffering of packets. If the service rate is poor, the queue builds up at a very faster
rate, degrading the performance of the network. Lastly, it is a measure of how effective a system
is in reducing the idle durations of channels and how it improves the utilization of the channels.
In Fig 11, the effects of packet drop rate on system performance are analyzed in a network with 4
nodes. A system with shorter waiting times is desired as it suffers lower packet drop rates.
Fig 11. Impact of transmission delays in a network with 4 nodes
The transmission delays have a huge impact on the AMCP in Fig 11, and its packet drop rate is
higher than that of the proposed scheme. The higher drop rate shows that the scheme is more
prone to the MSC which result in the poor utilization of channels causing longer idle durations.
The CSA had a lower packet drop rate owing to its ability to address the effects of the MSC. Its
response time is superior resulting in lower packet drop rate, and shorter end-to-end delays.
In Fig 12, the CSA scheme also outperformed the AMCP. The network size was increased to 6
nodes to create a congested network state which is ideal for testing the scalability of protocols.
The results demonstrate that the CSA is more scalable and more efficient in addressing the effects
of the MSC.

14
Fig 12. The analysis of the effect of MSC in a network with 6 nodes
Lastly, the size of the network was increased to 30 nodes in Fig 13. The data flows were almost 4
times the number of the channels resulting in a heavily congested and over loaded network. The
CSA was superior and had lower packet drop rate in all but 2 nodes. It is superior, more scalable
and effective in addressing the effects of the MSC. It improves the utilization of the channels, the
system service rate, and the system response time furthermore, it also reduces the idle durations
of the channels and the end-to-end delay.
The performance of the two schemes is poor in the central nodes largely due to the fact that as the
inter arrival rate of packets increases, the buffers fill up at a faster rate while the congestion also
increases steadily. When these to occurrences take places simultaneously, the performance of the
network is degraded. However, performance thereafter improves with the passage of time when
the schemes take advantage of spatial reuse. The TCP protocol also adjusts its transmission
window when it detects congestion resulting in fewer packets being dropped. The superiority of
the CSA is demonstrated by the fact that its drop rate is lower packets in comparison with that of
the reference model.
Fig 13. Evaluation of the CSA in a congested network

15
8.0 Conclusion
The proposed CSA protocol reduces the MSC and limits it to the first cycle. It also reduces the
idle durations of the channels and the signaling overhead cost. The system service rate and system
response time are also improved which in turn improves the overall system performance. The
results demonstrate that the CSA performs better in large and congested networks, which
illustrate the scalability and the effectiveness properties of the scheme. In this case, a large
network has 10 or more nodes. A distinguishable pattern was observed in large networks. The
performance of the network worsens in the middle nodes before it improves again in the last
nodes. This is caused by congestion and an increase in the data traffic however; it improves in the
last nodes due to spatial reuse and the adjustment of the TCP window.
References
[1] R Maheshwari, H Gupta and S R. Das. “Multichannel MAC protocols for wireless networks”.
SECON 2006, Volume 2, Issue , 28-28 Sept. 2006.
[2] J So and N Vaidya (2004). “Multichannel MAC for Ad Hoc Networks: Handling Multichannel
Hidden Terminals Using A Single Transceiver”. MobiHoc 2004, Roppongi, Japan, ACM.
[3] J Chen and S Sheu. “Distributed Multichannel MAC protocol for IEEE 802.11 ad hoc wireless
LANs”. Computer Communications (2005). www.elsevier.com/locate/comcom
[4] H W So and J Walrand (2006). “Design of a Multichannel Medium Access Control Protocol for Ad-
Hoc Wireless Networks”. University of California at Berkeley, EECS Department.
http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-54.html
[5] W Chen, J Liu, T Huang, and Y Chang (2008). “TAMMAC: An Adaptive Multichannel MAC
Protocol for MANETs”. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 7,
NO. 11, NOVEMBER 2008
[6] A Muir and J.J. G Aceves.“A channel access protocol for multi hop wireless networks with multiple
channels. University of California.”IEEE International Conference on Communications, 07 Jun 1998
–11 Jun 1998, Atlanta, GA , USA.
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=683102&tag=1
[7] T Luo, M Motani, and V Srinivasan (2009). “Cooperative Asynchronous Multichannel MAC: Design,
Analysis, and Implementation”. IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 8, NO.
3, MARCH 2009
[8] A K Mahani, M Naderi, C Casetti, and C F. Chiasserini (2007). “ENHANCING CHANNEL
UTILIZATION IN MESH NETWORKS”. IEEE MILCOM, Orlando, FL, 2007, 29-31 October 2007.
www.ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4454826
[9] J Nitin, S R. Das and A Nasipuri. “A multichannel CSMA MAC protocol with Receiver-Based
Channel Selection for Multihop Wireless Networks.” Proceedings. Tenth International Conference on
Computer Communications and Networks, 2001. 15 Oct 2001 - 17 Oct 2001, Scottsdale, AZ , USA.
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=956301
[10] J Shi, T Salonidis, and E W. Knightly (2006). “Starvation Mitigation Through Multichannel
Coordination in CSMA Multi-hop Wireless Networks”, MobiHoc’06, May 22–25, 2006, Florence,
Italy
[11] R Zhao and B Walke (2005). “Decentrally controlled wireless multi hop mesh networks for high
quality multi media communications.” MSWiM October 2005.
http://portal.acm.org/citation.cfm?id=1089444.1089479 6 June 2007
[12] M Razfar and A Abedi (2011). “Single Channel Versus Multichannel MAC Protocols for Mobile Ad
Hoc Networks.” Proceedings of the World Congress on Engineering and Computer Science 2011 Vol
II, WCECS 2011, October 19-21, 2011, San Francisco, USA
[13] C Shi1, H Zhao1, Y Tang, and Y Zhou (2012). “Analysis of the Multi-channel MAC Protocols in
Wireless Networks.” Proceedings of the 2012 2nd International Conference on Computer and
Information Application (ICCIA 2012)
[14] S. M. KAMRUZZAMAN, Md. Abdul HAMID, M. Abdullah-Al-WADUD (2010). “An Energy-
Efficient MAC Protocol for QoS Provisioning in Cognitive Radio Ad Hoc Networks.”
RADIOENGINEERING, VOL. 19, NO. 4, DECEMBER 2010

16
[15] S. M. Kamruzzaman (2010). “An Energy Efficient Multichannel MAC Protocol for Cognitive Radio
Ad Hoc Networks.” International Journal of Communication Networks and Information Security
(IJCNIS) Vol. 2, No. 2, August 2010
[16] H A Omar, W Zhuang and L Li (2011). “VeMAC: A Novel Multichannel MAC Protocol for
Vehicular Ad Hoc Networks.” IEEE MobiWorld 2011
[17] N Trivedi, G. Kuamr, T Raikwar (2013). “Survey on MAC Protocol for Wireless Sensor Network.”
International Journal of Emerging Technology and Advanced Engineering, Volume 3, Issue 2,
February 2013
[18] L Tang, Y Sun, O Gurewitz, and D B. Johnson (2011). “EM-MAC: A Dynamic Multichannel Energy-
Efficient MAC Protocol for Wireless Sensor Networks.” MobiHoc’11, May 16–19, 2011, Paris,
France
[19] Md. A Hamid, M. Abdullah-Al-Wadud, and I Chong (2010). “A Schedule-based Multi-channel MAC
Protocol for Wireless Sensor Networks.” Sensors 2010, 10, 9466-9480; doi:10.3390/s101009466,
www.mdpi.com/journal/sensors, ISSN 1424-8220
[20] V Kumar and J Thakur (2013). “Multi-Channel Mac Protocols in Wireless Sensor Networks: A
Review.” International Journal of Advanced Research in Computer Science and Software
Engineering, Volume 3, Issue 10, October 2013, ISSN: 2277 128X
[21] S Hsu, C Hsu, S Lin, and F Lin (2010). “The Analysis of a Self-stabilizing MAC Protocol for Multi-
channel Ad Hoc Networks.” JOURNAL OF NETWORKS, VOL. 5, NO. 8, AUGUST 2010
[22] K H. Almotairi and X (Sherman) Shen (2011). “Multichannel medium access control for ad hoc
wireless networks.” WIRELESS COMMUNICATIONS AND MOBILE COMPUTING Wirel.
Commun. Mob. Comput. (2011)
[23] T Luo, M Motani, V Srinivasan.” CAM-MAC: A Cooperative Asynchronous Multi-Channel MAC
Protocol for Ad Hoc Networks” Broadband Communications, Networks and Systems, 2006.
BROADNETS 2006
[24] J Nieminen and R Jäntti (2011). “Delay-throughput analysis of multi-channel MAC protocols in ad
hoc networks.” EURASIP Journal on Wireless Communications and Networking 2011
[25] B. S Kumar, P. SivaChandrika, A. S Kumar, and K. Srinivasulu (2013). “Energy - Effectual
Strategies for Conjunctive Multichannel MAC Protocols.” International Journal of Advanced
Research in Computer and Communication Engineering Vol. 2, Issue 2, February 2013
[26] M Velempini and M. E. Dlodlo (2009). “Analyzing the effects of increasing data channels and the
number of data flows on network performance.” Wireless VITAE 2009, Aalborg, Denmark, 17-20
May, 2009.
[27] M Velempini and M. E. Dlodlo (2009). “A multiple channel selection and coordination MAC
Scheme.” Mesh 2009, Glyfada/Anthens, Greece, 18 – 23 June 2009.
[28] E Jung and N H. Vaidya (2002). “A Power Control MAC Protocol for Ad Hoc Networks.”
MOBICOM’02, September 23–28, 2002, Atlanta, Georgia, USA.
[29] The editors of IEEE 802.11. Wireless LAN Medium Access Control (MAC) and Physical Layer
(PHY) Specification, 1997.

Addressing the Multichannel Selection, Scheduling and Coordination Cost

More Related Content

Similar to Addressing the Multichannel Selection, Scheduling and Coordination Cost (20)

More from pijans (20)

Recently uploaded (20)

Addressing the Multichannel Selection, Scheduling and Coordination Cost