International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
Parameter Based Taxonomy for Channel Allocation
in MRMC WMN
MRUTYUNJAY G. HIREMATH , NITYA N. KULKARNI Y, JAYALAKSHMI G. NARAGUND Z AND
R.M. BANAKARX
y
PG Student, Department of Computer Science and Engineering, BVBCET, Hubli, India
PG Student, Department of Computer Science and Engineering, BVBCET, Hubli, India
z
Professor, Department of Computer Science and Engineering, BVBCET, Hubli, India
x
Professor,Department of Electronics and Communication Engineering, BVBCET, Hubli, India
Abstract—Wireless Mesh Networks are emerging trend in
Wireless Networks for their low cost, easy access, self healing and
self organizing capabilities. Despite the fact that WMNs work
efficiently, they are heavily impacted by interference in the channel.
To overcome this problem several researches suggest Multiple
Radios and Multiple Channels (MRMC) in WMN. MRMC networks
improve the overall capacity and throughput of the network. The
MRMC WMN includes many issues like Channel Allocation (CA),
scheduling and routing. Lot of research work is done and is going on
to solve the issues of CA in MRMC WMNs. The authors of this
article aim to propose the taxonomy of different CA algorithms
based on key performance parameters namely throughput, network
capacity and both. Network throughput is the rate of successful
message delivery, so as throughput increases, distribution of message
between the users also improves. Scalability of network enhances the
network capacity to fulfill high demand of internet connection in the
world. The scalability of MRMC WMN is analyzed through
simulation using network simulator NS3.22. The results of the
simulation shows that the selections of efficient CA scheme plays an
important role to improve the network performance.
Index Terms—Graph algorithms, Multi radio and multi channel communication, Channel allocation, WMN.
I. INTRODUCTION
Wireless Mesh Network (WMN) is a reliable, multi-hop, high
bandwidth and inexpensive broadband access network. It consists of
mesh routers, mesh clients and gateways. Mesh routers form the
backbone of the network providing com-munication to mesh clients
among themselves and to/from the external network. Gateways
provide internet access to all the nodes of WMN through wired
connection. Mesh clients are laptops, mobile devices and other
wireless devices that perform simple client functionalities. The
capability of mesh routers to forward packets along with the basic
access point tasks appeals various applications. Some of the areas
where WMNs find applications in are security and surveillance
systems, building automation, health and medical systems,
transportation systems, community networks, city networks etc.
WMNs are utilized so widely because of their various benefits such
as increased reliability, low installation cost, large coverage area and
automatic network connectivity. De-spite the fact that WMNs are
gaining popularity, there exists some issues such as stability in
topology, multi channel, routing, interference etc. which need to be
addressed with
innovative methodologies to improve throughput and delay of
the network.
One of the challenges to enhance the throughput is to cope with
the interference in the network. The interference in the transmission
on same frequency channels can be eliminated by utilizing multiple
radios with different channels assigned to them. To utilize this
environment of Multiple Radio Multiple Channel (MRMC) in WMN
efficiently, the channels must be assigned, scheduled and routed
effectively within network capacity while maintaining network
connectivity. Although WMN provides simultaneous transmission by
using multiple radios and multiple channels, interference cannot be
excluded completely. The interference mainly depends on the
channels assigned to nodes which is carried out by CA schemes.
Hence the choice of channel assignment plays an important role in
MRMC environment.
In this paper various channel assignment schemes are studied
and reviewed. The authors focus on presenting the taxonomy for
different CA mechanisms based on the param-eters such as
throughput and network capacity. In an effort to visualize the
MRMC environment of WMN, the authors also concentrate on
the performance analysis of this network using network simulator
NS3.
The paper is organized in the following sections; section II
presents literature survey on various channel assignment
algorithms. Section III provides MRMC WMN architecture,
its importance and classification of CA based on parameters.
Section IV discusses the result analysis of MRMC implementation in NS3. Finally the authors conclude the paper in
section V.
II. LITERATURE REVIEW
Channel Assignment is one of the widely explored area of
WMNs. Several research have been conducted and new
methodologies are proposed to improve different performance
parameters of mesh networks. Some of the approaches focus on
orthogonal channels and some on both orthogonal and partially
overlapped channels. The literature survey conducted in this
paper is related to the various channel assignment schemes
recently proposed.
Since the number of orthogonal channels are few, two nodes
might employ same frequency channel for data transmission. This
leads to co-channel interference, elimination of which is
283
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
the main focus of the authors of [1]. They propose I-Matrix
interference model that helps in selecting channels with less
interference, based on POC channel assignment algorithm. The Imatrix generated consists of the interference factor for all channel
combination of radios. The channel combination with least
interference factor is used for assignment. The algorithm also
increases the number of simultaneous transmissions and eventually
upgrades the network capacity.
Sok-Hyong Kim and Young-Joo Suh [2] proposed a Rate-Based
Channel Assignment (RB-CA) algorithm to alleviate performance
anomaly in multi-channel multi-rate WMNs. By exploiting multiple
channels, the proposed algorithm alters a low-rate single-hop path to
a high-rate multi-hop path. Thus, the algorithm delivers a large
volume of traffics in WMNs simultaneously from or to the Internet
through multiple non-overlapping channels along with high-rate
links.
The classic cluster based channel assignment scheme (CCAS)
was proposed in [3]. The author exploits multiple paths between
mesh router and gateway to or form clusters. The clusters formed
are assigned channels to minimize co-ordinated and non
coordinated interference, thereby increasing the capacity of
individual links.
Vinay kapase [4] proposed a Cluster based Channel As-signment
algorithm which works in three phases. The first phase involves
calculation of Euclidean distance between two wireless mesh nodes
and the formation of clusters. In the second phase interface allocation
and channel assignment using greedy technique is carried out. In the
third phase Channel reassignment occurs considering interference
from nearby nodes operating on the same channel.
The greedy technique is also utilized in the distributed
channel assignment algorithm as proposed by Ko et al. in [5].
The technique minimizes channel’s local objective function
depending only on local information of nodes. Every node
selects a channel that minimizes the sum of interference cost
within its interference range.
Yan Xia [6] proposed channel assignment based on selforganized grid-loop i.e. forming grid-loops via Minimum
Spanning Tree and through group channel. Differing from classic
cluster topology, the construction of grid-loop topology not only
has an efficient constructing mechanism, but also has a simple
and robust structure. In each grid loop every node is assigned
with an identifier(ID) . The mesh node with larger ID becomes
loop creator and assigns channels for self radios. Later it assigns
available channels to the remaining nodes in the loop.
The algorithm proposed by aizaz and nazia [7], TICA uses
topology control algorithm based on power control to assign
channels to multi-radio mesh routers. It helps to minimize co-channel
interference, maximize network throughput and maintain network
connectivity. Topology control algorithm selects the node with
minimum transmission power and which is closer to other nodes. The
node closer to gateway with less transmission power is assigned the
highest rank. The channels are assigned to nodes in the decreasing
order of their rank.
direction of radio antennas. Considering this fact the authors of [8]
utilize directional antennas to assign channels for multi radio WMN.
The proposed mechanism involves construction of the auxiliary
network graph. The number of radios on each node and the total
number of available non-overlapping channels will be formulated
from this graph. There are certain constraints that need to be adhered
such as dual path routing should be conducted and access points
should be one-to-one associated with the gateway nodes. The flow
rate of each AP-GN(Access Point-Gateway Node) pair on each link
determines the routers and bandwidth allocated to each AP. This
method drastically improves aggregate system throughput while
maintaining fair bandwidth allocations among APs.
Game theory is a mathematical tool, particularly useful, in the
network engineering field to model highly complex scenarios. [9]
utilizes this concept to design a systematic approach to minimize
adjacent channel interference by con-sidering partially
overlapped channels in WMNs. The authors proposed the
distributed channel assignment algorithm which improves the
overall performance of network by enhancing throughput and
channel reuse.
Previously several researchers surveyed the various channel
assignment schemes identifying the issues and challenges. The
authors of [10] identified key challenges and research approaches in
CA for multiple radios, infrastructure, topology , graph theory and
routing dependency. They also concentrate on the goals and
objectives of an efficient algorithm and provide classification and
comparative analysis of different schemes.
The literature survey conducted in this section helps to
identify the key parameters of channel assignment algorithms
that have great deal of impact on network performance. The
most focused parameters are network throughput and network
capacity. Hence the authors are motivated to present the
taxonomy of CA schemes based on these parameters as
discussed in section IV.
III. MR-MC ARCHITECTURE
Wireless mesh network as discussed in section I consists of mesh
router, mesh clients and mesh gateway as nodes. Depending on the
functionality of these nodes, the architecture of WMN can be
classified as infrastructure/backbone, client WMN and hybrid WMN
[11]. The infrastructure/backbone WMN consists of backbone
network formed by mesh routers and all the traffic are routed by
them. Client WMNs are peer to peer networks wherein the client
nodes do not use mesh routers to communicate and perform
configuration and routing. The hybrid WMN architecture employs
both infrastructure and client meshing in the WMN network, thus
utilizing the merits of both architectures.
Figure 1 shows the infrastructure based WMN, mesh routers and
gateway, forming the backbone of the network. Clients can
communicate with the help of routers in the backbone network. For
the communication to occur between mesh router and client, both the
nodes should have same radio technology otherwise the clients have
to connect to the base station of
The interference among the nodes is also effected by the
284
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
Fig. 1. MRMC WMN Architecture
the network. The radio technology could belong to 802.11 or other
radios as well. The capacity of such communication in networks
could be enhanced by either using efficient MAC protocol for single
channel nodes or by utilizing multiple channels for nodes. Further
multiple radios could be assigned to nodes with multiple channels to
exploit
simultaneous
trans-missions,
improving
network
performance.
In MRMC WMN, each mesh router is equipped with multiple
radios, each allocated to non-overlapping channels of the radio
frequency spectrum as shown in figure 1. The links between the
nodes are alloted different colors to emphasize on the multiple
channels used by the nodes. Mesh routers can communicate to
multiple other routers simultaneously as it owns multiple radios. A
virtual MAC protocol such as the multi-radio unification protocol
(MUP) [12] is required above the MAC to coordinate
communications in all channels. For simplicity of design and
application, in this paper a single channel is used in each radio.
Regarding the access to the medium, mesh routers implement the
mesh coordination function (MCF). As is described in [13], MCF
consists of a mandatory and an optional scheme. For the mandatory
part, MCF relies on the contention-based protocol known as Enhanced Distributed Channel Access (EDCA), which itself is an
improved variant of the basic 802.11 distributed coordination
function (DCF). Using EDCA, a router may transmit multiple frames
whose total transmission duration may not exceed limit. To enhance
QoS, MCF describes an optional medium access protocol called
Mesh Coordinated Channel Access (MCCA). It is a distributed
reservation protocol that allows Mesh nodes to avoid frame
collisions. After an MCCA transmission ends, Mesh routers use
EDCA for medium contention again [14].
The routing in WMN is conducted in the MAC layer using Hybrid
Wireless Mesh Protocol (HWMP) and Peer Link Man-
agement Protocol(PLM). The Peer Link Management Protocol is
applied when new mesh routers enter or exit the existing WMN. The
protocol avoids the collisions in the beacon frames transmitted by the
mesh routers. It establishes link between peers. The path selection in
the network for data transmission is carried out by HWMP. The
protocol uses the radio aware routing metric Airtime Link Cost
Metric(ALM) which evalu-ates the link quality by measuring channel
access and routing overhead along with frame error rate. The
protocol works in two modes depending on the stability of the
network topology, proactive and reactive mode. In both the modes,
nodes transmit the control packets namely, PREQ( Path Request),
PREP (Path Reply) and PERR (Path Error).
In the proactive PREQ mode, the root node periodically
broadcasts the PREQ packets. The mesh router receiving this PREQ
packet updates itself with the information of root node and the
distance between them and then forwards the packet. The router
sends the gratuitous PREP packet as a response to the root node if the
PREP bit is set to 0 in PREQ packet. In the proactive RANN (Root
Announcement) mechanism, the root node periodically floods RANN
packet into the network. The mesh router which receives the RANN
packet if needs to create/refresh route to the root then it sends a
unicast PREQ to the root. When the root receives this unicast PREQ,
it responds with a PREP to the Mesh router.
The reactive mode is based on the AODV protocol as specified in
RFC 3561 [15]. In the reactive mode, the source mesh router
transmits the PREQ packet in order to find the path to the destination
node. The intermediate nodes that receive the PREQ packet, update
the path to the source node if the sequence number in the packet is
greater than the previous one or the metric is better. If the
intermediate node does not have the path to the destination it simply
forwards the PREQ packet otherwise the action of the node depends
on the DO(Destination Only) and RF(Reply- and -Forward) flags. If
only the DO flag is set, the intermediate nodes forward the PREQ
packet until it reaches the destination node. The destination node in
turn replies with the PREP packet, allowing the intermediate nodes to
learn the path to the destination. If both the flags are not set and the
intermediate node is aware of the path to the destination node then it
transmits unicast PREP packet to the source router and does not
forward the PREQ packet. In the case where DO flag is unset and RF
flag is set, the intermediate node sends the unicast PREP packet to
the source node and sets the DO flag while forwarding the PREQ
packet. This leads to subsequent intermediate nodes not sending the
unicast PREP packet back to the source node.
As discussed before, network capacity could be boosted
through the usage of multiple channels for nodes. In this view,
the selection of Channels for assignment to nodes play a
significant role. Hence various CAAs are reviewed and
classified in section IV.
IV. PROPOSED WORK
Channel Assignment is one of the fundamental aspect of
MRMC Wireless Mesh Network which has to be dealt
285
International Journal of Emerging Technology in Computer Science &
Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
TABLE I
TAXONOMY BASED ON PARAMETERS
Parameter
Throughput
Methodology
centralized channel assignment and routing
scheme based on heuristic route discovery and
traffic load estimation[8]
Clustered based channel assignment [16]
Load model CA [17]
Extended centralized traffic independent algorithm and MSTCA ( Minimum Spanning Tree
CA) [4]
Loop based dynamic CA [6]
CA using network coding and using centralized,
auction-based algorithm [18]
A game theoretic distributed CAA [9]
Network Capacity
Throughput and
Network Capacity
Partially overlapped channel assignment algorithm [19]
Imatrix interference model [1]
Data rate based CA [2]
Topology-controlled Interference-aware Channel
assignment Algorithm [7]
CRAFT (Channel and Routing Assignment with
Flow Traffic) [5]
Optimization model and CA based on decision
variables [20]
Multi-Channel Minimal Number of transmissions Trees [21]
Mixed integer linear programming model [22]
Multi objective algorithm for channel assignment[23]
with precision and thought. In this regard, several research has been
done and various CA scheme, improving different parameters of the
network are proposed. These CA schemes are designed to satisfy
certain objectives such as improvement in overall throughput of the
network, improvement in total network capacity, minimizing the
average delay in the net-work, elimination or rather minimization of
interference in the network etc. The authors concentrate on two such
parameters; throughput and network capacity of CA schemes based
on which the CA algorithms are classified as shown in table I.
HWMP
ALM ( Airtime Link cost Metric)
ALM
ALM
ALM
Optimized Link State Routing
Protocol (OLSR)
Optimized Link State Routing
Protocol (OLSR)
ALM
ALM
AODV (Ad-hoc On-Demand
Distance Vector) routing protocol
ALM
ALM
Multicast routing metric
MR (Multi Radio)-LQSR
(Link Quality Source ROuting)
Multicast routing
TABLE II
UNDERLYING ENVIRONMENT SETUP
SPECIFICATION
Underlying
framework
Propagation
Delay
model
Propagation
loss
model
Error Rate Model
Bandwidth
Starting frequency
RTS/CTS
Remote Station Manager
step size
Random Start
Simulation Time
Packet Interval
Packet Size
Number of interfaces
Root
V. RESULT ANALYSIS
One of the most widely applied area of WMN is community
networking. Such networks involve rapid growth in the number
of users as the community scales. Hence the network should
support scalability. To visualize such environment, authors focus
on analyzing the behavior of MRMC WMN with the increase in
number nodes. The simulation of MRMC WMN is conducted
using network simulator NS3.22. The performance study is
carried out with respect to throughput and delay. The underlying
environment set for the simulation is as shown in table II.
The nodes are arranged in grid topology, each with two
radios. The radios are assigned non-overlapping channels
statically. The distance between two edge nodes is 50m. The
simulation is initiated randomly at 0.1s and is conducted for
200s.
Routing
Dual path routing protocol
VALUE
IEEE 802.11a
ConstantSpeedPropagationDelayModel
LogDistancePropagationLossModel
NistErrorRateModel
5Mbps
5 GHz
Disabled
ArfWifiManager
50 m
0.1s
200s
0.1s
1024 bytes
2
MAC address of a node in proactive mode and ff:ff:ff:ff:ff:ff in reactive
mode
It can be interpreted from figure 2 and figure 3 that as the number of
nodes increases throughput decreases and delay increases respectively.
Since the CA scheme used is static with number of channels same as that
of number of radios for each node, the behavior of both the curves is
simplistic. With the
286
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
[2] S.-H. Kim and Y.-J. Suh, “Rate-based channel assignment algorithm for
multi-channel multi-rate wireless mesh networks,” in Global Telecommunications Conference, 2008. IEEE GLOBECOM 2008. IEEE. IEEE,
2008, pp. 1–5.
[3] A. Naveed and S. S. Kanhere, “Cluster-based channel assignment in multiradio multi-channel wireless mesh networks,” in Local Computer Networks,
2009. LCN 2009. IEEE 34th Conference on. IEEE, 2009,
pp. 53–60.
[4] V. Kapse and U. Shrawankar, “Interference-aware channel assignment
for maximizing throughput in wmn,” arXiv preprint arXiv:1305.2838,
2013.
[5] B.-J. Ko, V. Misra, J. Padhye, and D. Rubenstein, “Distributed channel
assignment in multi-radio 802.11 mesh networks,” in Wireless Communications and Networking Conference, 2007. WCNC 2007. IEEE. IEEE,
2007, pp. 3978–3983.
[6] Y. Xia, Y. Zeng, B. Fu, and Z. Gong, “Loop-based dynamic channel
assignment for multi-radio multi-channel wireless mesh networks,” in
Fig. 2. Number of Nodes vs Throughput
Networks Security Wireless Communications and Trusted Computing
(NSWCTC), 2010 Second International Conference on, vol. 1. IEEE,
2010, pp. 265–268.
[7] A. U. Chaudhry, R. H. Hafez, O. Aboul-Magd, and S. A. Mahmoud, “Faulttolerant and scalable channel assignment for multi-radio multi-channel ieee 802.11
a-based wireless mesh networks,” in GLOBECOM Workshops (GC Wkshps),
2010 IEEE. IEEE, 2010, pp. 1113–1117.
[8] W. Zhou, X. Chen, and D. Qiao, “Practical routing and channel assignment
scheme for mesh networks with directional antennas,” in
Communications, 2008. ICC’08. IEEE International Conference on.
IEEE, 2008, pp. 3181–3187.
[9] P. B. Duarte, Z. M. Fadlullah, A. V. Vasilakos, and N. Kato, “On the
partially overlapped channel assignment on wireless mesh network backbone: A game theoretic approach,” Selected Areas in Communications, IEEE
Journal on, vol. 30, no. 1, pp. 119–127, 2012.
[10] M. A. Hoque and X. Hong, “Channel assignment algorithms for mrmc
wireless mesh networks,” International Journal of Wireless & Mobile
Networks, vol. 3, no. 5, pp. 76–94, 2011.
[11] I. F. Akyildiz, X. Wang, and W. Wang, “Wireless mesh networks: a
survey,” Computer networks, vol. 47, no. 4, pp. 445–487, 2005.
[12] A. Adya, P. Bahl, J. Padhye, A. Wolman, and L. Zhou, “A multi-radio
unification protocol for ieee 802.11 wireless networks,” in Broadband
Networks, 2004. BroadNets 2004. Proceedings. First International Conference on. IEEE, 2004, pp. 344–354.
Fig. 3. Number of Nodes vs Delay
utilization of efficient CA scheme exploiting more number of
channels per radio of a node, improvement in throughput and
delay could be achieved.
VI. CONCLUSION
In this paper we briefly discussed the architecture and vari-ous
channel assignment algorithms(CAA) of MRMC WMNs. The CA
schemes are classified based on the parameters throughput and
network capacity. The classification is pre-sented in the form of
taxonomy. For simplicity MRMC en-vironment of WMN is
simulated using NS3.22 with static channel assignment. The
performance analysis of the sim-ulation is conducted with respect to
scalability metric. The analysis shows that with the utilization of
efficient CA scheme, the throughput and delay could be improved
with respect to behavior of metrics presented in this paper. The paper
also helps other researchers to select the CA scheme based on the
parameters.
REFERENCES
[1] M. A. Hoque, X. Hong, and F. Afroz, “Multiple radio channel assignement utilizing partially overlapped channels,” in Global Telecommunications Conference, 2009. GLOBECOM 2009. IEEE. IEEE, 2009, pp.
1–7.
[13] G. R. Hiertz, D. Denteneer, S. Max, R. Taori, J. Cardona, L. Berlemann, and
B. Walke, “Ieee 802.11 s: the wlan mesh standard,” Wireless
Communications, IEEE, vol. 17, no. 1, pp. 104–111, 2010.
[14] N. del Projecte, “Projecte final de carrera.”
[15] C. Perkins, E. Belding-Royer, S. Das et al., “Rfc 3561-ad hoc on-demand distance
vector (aodv) routing,” Internet RFCs, pp. 1–38, 2003.
[16] S. Max, Y. Zang, Y. Zhou, and H. Zhang, “Model-based radio and channel assignment in ieee 802.11 wireless mesh networks,” in
GLOBECOM Workshops, 2009 IEEE. IEEE, 2009, pp. 1–6.
[17] K. Athota, A. Negi, and C. R. Rao, “Interference-traffic aware chan-nel
assignment for mrmc wmns,” in Advance Computing Conference (IACC),
2010 IEEE 2nd International. IEEE, 2010, pp. 273–278.
[18] S. Chieochan and E. Hossain, “Channel assignment for throughput
optimization in multichannel multiradio wireless mesh networks using
network coding,” Mobile Computing, IEEE Transactions on, vol. 12,
no. 1, pp. 118–135, 2013.
[19] D. Wang, P. Lv, Y. Chen, and M. Xu, “Pocam: Partially overlapped chan-nel
assignment in multi-radio multi-channel wireless mesh network,” in
Communications and Information Technologies (ISCIT), 2011 11th
International Symposium on. IEEE, 2011, pp. 188–193.
[20] S. Shah, H. Hussain, and M. Shoaib, “Minimizing non-coordinated interference in multi-radio multi-channel wireless mesh networks (mrmcwmns),” in Digital Information Management (ICDIM), 2013 Eighth
International Conference on. IEEE, 2013, pp. 24–28.
[21] H. L. Nguyen and U. T. Nguyen, “High-performance multicast routing in multichannel multi-radio wireless mesh networks,” in Global Telecom-munications
Conference (GLOBECOM 2010), 2010 IEEE. IEEE, 2010,
pp. 1–6.
[22] C.-J. Lin, S.-H. Lin, and C.-F. Chou, “Performance study of optimal
routing and channel assignment in wireless mesh networks,” in Global
Telecommunications Conference, 2007. GLOBECOM’07. IEEE. IEEE,
2007, pp. 4818–4822.
287
International Journal of Emerging Technology in
Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 14 Issue 2 –APRIL 2015.
[23] M. Sharifi Jebeli and M. Dehghan, “Joint multicast routing and channel
assignment in multiradio multichannel wireless mesh networks using a multi
objective algorithm,” in Information and Knowledge Technology (IKT), 2014 6th
Conference on. IEEE, 2014, pp. 163–170.
288