Routing in Wireless Mesh Networks: Two Soft Computing Based Approaches

International Journal of Mobile Network Communications & Telematics ( IJMNCT) Vol. 3, No.3, June 2013
DOI : 10.5121/ijmnct.2013.3304 29
Routing in Wireless Mesh Networks: Two Soft
Computing Based Approaches
Sharad Sharma1,
Shakti Kumar2
and Brahmjit Singh3
1,3
Deptt. of Electronics & Communication Engineering, National Institute of Technology,
Kurukshetra, India
2
Computational Intelligence (CI) Lab, IST Klawad, Yamunanagar, India
1
sharadpr123@rediffmail.com , 2
shaktik@gmail.com , 3
brahmjit.s@gmail.com
ABSTRACT
Due to dynamic network conditions, routing is the most critical part in WMNs and needs to be optimised.
The routing strategies developed for WMNs must be efficient to make it an operationally self configurable
network. Thus we need to resort to near shortest path evaluation. This lays down the requirement of some
soft computing approaches such that a near shortest path is available in an affordable computing time. This
paper proposes a Fuzzy Logic based integrated cost measure in terms of delay, throughput and jitter.
Based upon this distance (cost) between two adjacent nodes we evaluate minimal shortest path that updates
routing tables. We apply two recent soft computing approaches namely Big Bang Big Crunch (BB-BC) and
Biogeography Based Optimization (BBO) approaches to enumerate shortest or near short paths. BB-BC
theory is related with the evolution of the universe whereas BBO is inspired by dynamical equilibrium in
the number of species on an island. Both the algorithms have low computational time and high convergence
speed. Simulation results show that the proposed routing algorithms find the optimal shortest path taking
into account three most important parameters of network dynamics. It has been further observed that for
the shortest path problem BB-BC outperforms BBO in terms of speed and percent error between the
evaluated minimal path and the actual shortest path.
KEYWORDS
Wireless Mesh Network, Cost function, Fuzzy Logic, Big Bang Big Crunch, Biogeography Based
Optimization
1. INTRODUCTION
Wireless Mesh Networks are rapidly deployable, dynamically self organizing, self configuring,
self healing, self balancing and self aware multi hop networks. In these networks each node
(stationary or mobile) has the capability to join and create a network automatically by sensing
nodes with a similar capability within its radio range. WMNs can be categorized to three types:
(1) Infrastructure Mesh; (2) Client Mesh and (3) Hybrid Mesh [1].
The routing algorithms of a WMN must work in a decentralized, self-organizing and self
configuring manner. The routing protocols developed for MANETs can usually be applied to
WMNs as well e.g. Ad hoc On-demand Distance Vector (AODV) [2], Topology Broadcast based
Reverse Path Forwarding (TBRPF) [3], Dynamic Source Routing (DSR) [4] etc. The existing
MAC and routing protocols applied to WMNs do not provide enough scalability [5]. The factors
like resource allocation, interference avoidance and rate adaptation across multiple hops critically
affects the routing path selection [6]. In a WMN the performance parameters can be categorized

30
as per flow; per node; per link; inter flow and network wide parameters. These routing metrics are
Airtime Cost Routing Metric [6], Hop Count, Per-Hop Round Trip Time (RTT) [7], Metric of
Interference and Channel-Switching (MIC) [8], Expected Transmission Count (ETX) [9],
Expected Transmission on a Path (ETOP) [10], Expected Transmission Time (ETT) and
Weighted Cumulative ETT (WCETT) [11], Low Overhead Routing Metric [12, 13], Effective
Number of Transmissions (ENT) [14], Bottleneck Link Capacity (BLC) [15], Per-Hop Packet
Pair Delay and Expected Data Rate (EDR) [16], etc. The comparison of performance metrics on a
routing algorithm is discussed in [16]. Soft computing based techniques provide the optimal
solution and quickly adapts to dynamic environmental changes [17]. In this paper, we propose
two soft computing approaches for addressing the routing problem in WMNs, namely BB-BC and
BBO.
This paper is organized into five sections. Section I presents the motivation for the present work.
Section II introduces the Big Bang Big Crunch (BB-BC) and Biogeography Based Optimization
(BBO) algorithms. In section III we present a fuzzy logic based approach to evaluate integrated
link cost (ILC) for each link consisting of throughput, end-to-end delay and average jitter of a
given path. In section IV, BB-BC and BBO are employed to evaluate the shortest path in the
configured network. The section V compares the performance of two algorithms. Conclusions are
drawn in Section VI.
2. BIG BANG BIG CRUNCH (BB-BC) AND BIOGEOGRAPHY BASED
OPTIMIZATION (BBO)
2.1 BB-BC Optimization Algorithm
The Big Bang theory is one of the most widely accepted theories of the evolution of this universe
[18]. The BB-BC theory believes that energy discharged by the initial explosion i.e., kinetic
energy, is counterbalanced by the energy of bodies attraction known as gravitational pull. If there
is enough mass so that the later is bigger than the first when a critic density is reached, the
expansion will stop and the universe will start to contract, leading to an end very similar to its
beginning, named by the scientists as the Big Crunch (Great Implosion). In the Big Bang phase,
energy dissipation produces disorder and randomness as the main feature of this phase. In the Big
Crunch phase, randomly distributed particles are drawn into an order. This theory of repeated big
bang followed by big crunch phases forms the basis of an optimization algorithm called the Big
Bang-Big Crunch optimization algorithm [19, 20].
Primarily a set of candidate solutions (population) is generated randomly in the search space. The
fitness as defined by the objective function, of each solution is enumerated and ranked
accordingly. After the random Big Bang phase contraction is applied in Big Crunch phase to
compute the centre of mass as:
∑
=
∑
=
=
N
i
if
N
i
i
x
if
x C
1
1
1
1
(1)
where xc = position of the centre of mass; xi = position of ith candidate; f i
= fitness function
value of candidate i; N = population size.

31
Best fit individual can also be considered as the centre of mass instead of using Equation (1)
alternatively. Generate new population around the centre of mass by adding or subtracting a
normal random number whose value decreases as the iterations elapse. This can be formalized as
klrxx C
new
/+= (2)
where xc stands for center of mass, l is the upper limit of the parameter, r is a normal random
number and k is the iteration step. Then new point xnew
is upper and lower bounded.
Pseudo Code for optimal path evaluation in WMNs using BB-BC based Algorithm is shown in
Figure 1.
begin
/ BB-BC Parameter Initialization for WMN/
Define Number and location of the nodes Source Node, Terminal Node, Number of
Paths, Number of Iterations,
/ End of BB-BC Parameter Initialization/
/Building of paths and Connectivity Matrix/
for i = 1 : n / all n Nodes /
for j = 1 : n / all n Nodes /
if distance (i, j) <= R (radio range of a node)
connectivity_matrix(i, j) = 1 /routing table maintenance/
Integrated_Link_Cost (i, j )= f (Throughput, Delay, Jitter)
/Integrated Link Cost Evaluation using Fuzzy System/
end if
end for j
end for i
/ Build paths between source and terminal node /
while (t < MaxGeneration or Termination Criteria not met)
Randomly generate initial population of k paths /Big Bang Phase/
Compute the ILC of all the candidate solutions
Sort the population from best to worst based on ILC /No.1 path is the Optimal path/
Compute the center of mass xc /Big Crunch Phase/
Generate new candidate solutions around xc by adding or subtracting a normal-
-random Number
end while
wait for stipulated time/ wait for an event
Postprocess results and visualization;
end
Figure 1: Pseudo Code of BB-BC based Algorithm for optimal path evaluation in WMNs
2.2 BBO ALGORITHM
BBO is the study of how species are articulated on the landscape in space and time. MacArthur
and Wilson (1963) [21] first suggested that the number of species of a given taxon that become
established on an island represents a dynamic equilibrium controlled by the rate of immigration of
new species and the rate of extinction (emigration) of previously established species. Based upon
the dynamical equilibrium theory Dan Simons [22] proposed BBO algorithm. Since its first
application, this meta-heuristic approach has been applied successfully to some engineering
applications. We apply this algorithm to evaluate minimal cost path.

32
BBO Algorithm [23, 24]:
• Initialize the parameters of BBO:
o maximum species count Smax and species count probability PS of each habitat,
maximum migration rates E and I , immigration rate λ, emigration rate µ, the
maximum mutation rate mmax , elitism parameter and habitat modification
probability
• Initialize a random set of habitats, each habitat corresponding to a potential solution to
the given problem.
• Do while not end of termination criteria (number of iterations)
• Compute “fitness” (HSI-habitat suitability index) for each habitat (solution). The
variables, characterizes habitability are Suitability Index Variables (SIVs).
• For each habitat, map the HSI to the number of species S. (Species Count is inversely
proportional to Cost)
• Compute λ and µ for each solution.
(3)
(4)
here the parameters are given as-
I- maximum possible immigration rate, E- maximum possible emigration rate, k- Number
of species of kth
individual and n- maximum number of species
• Modify each no elite habitat probabilistically using immigration and emigration rates &
recompute each HSI
• For each habitat, update the probability of its species count.
(5)
• Mutate each non-elite habitat based on its probability.
• If acceptable solution (optimal path) has been found then stop
• End
Pseudo code for optimum path selection in WMNs using BBO is presented in figure 2.
begin
/ BBO Parameter Initialization for WMN/
Define Number and location of the nodes, Source Node, Terminal Node, Number of
Paths, Number of Iterations, Smax , mmax , E, I, λ and µ
/Building of paths and Connectivity Matrix/
for i = 1 : n / all n Nodes /
for j = 1 : n / all n Nodes /
if distance (i, j) <= R (radio range of a node)
connectivity_matrix(i, j) = 1 /routing table maintenance/
Integrated_Link_Cost (i, j )= f (Throughput, Delay, Jitter)
/ILC Evaluation using Fuzzy System/
end if
end for j
end for i
/ Build paths between source and terminal node /
while (t < MaxGeneration or Termination Criteria not met)
Randomly generate initial population of k paths (Habitats) /Initial Population/
Compute the fitness (HSI) of all the candidate solutions (ILC)

33
Sort the population from best to worst based on ILC /No.1 path is the Optimal path/
Map the HSI to the no. of species of each individual
Compute the immigration rate λi and the emigration rate µi for each individual
Generate new candidate solutions and Update the probability
end while
wait for stipulated time/ wait for an event
Postprocess results and visualization;
end
Figure 2: Pseudo Code of BBO based Algorithm for optimal path evaluation in WMNs
3. System Model
To analyse and optimize the performance of routing algorithm of WMN environment simulations
were performed for a pre defined scenario in QualNet Simulator [25]. In this simulation, a
network of 25 nodes is considered that are placed within a 1500m X 1500m area and operating
over 300 seconds. A two ray ground propagation model is used with log normal shadow fading.
The transmission power of the nodes is set to 15dBm and the transmission range of the nodes is
250 meters. The data transmission rate is 2Mbits/sec. At the physical layer 802.11b and at MAC
layer MAC 802.11 is used. The traffic source is implemented using Constant Bit Rate (CBR),
sending at the rate of 1 packet/sec.
Figure: 3 Scenario of a 25 node static Wireless Mesh Network
20
14
9
19
7 8
18
12 13
17
6
2
5
2423
11
16
21
22
15
10
1 2 3 4 5

34
4. Path Cost Evaluation
For our model we considered 25, 64, 100 and 2500 nodes in this network. First node is assumed
to transmit data packets to the last node. To decide which path or route is to be used for any type
of traffic or condition depends on the current values of parameters at the nodes or the links.
Figure 4: Fuzzy System Model for Integrated Link Cost Evaluation
In the proposed work a routing metric of integrated throughput, end to end delay and jitter has
been considered at each node for all of its neighbouring nodes. Considering only forward links a
cost function has been proposed using fuzzy logic [26]. For our model to work we assume that
each node contains a software based fuzzy logic cost evaluation system as given in figure 4.
5. Results and Discussion
Two algorithms BB-BC and BBO were implemented in MATLAB v 7.6.0 (R2008a). Numerical
results were computed for both algorithms and are given in Table 1. It is observed from figure 5
and Table 1 that for a 25 node network, the minimum path cost is evaluated with zero error.
Further only 30 iterations suffice to make a quick decision about the shortest/near shortest
available path to ensure the optimal performance of the model. For 64 node WMN, initially the
BB-BC performs better (table 1). However as the number of iterations is increased from 30 to
100, the error of the two reduces to zero. Thus both again produced shortest path for the given
scenario. But time taken for finding the shortest path for BBO was 4.96 seconds as against 0.29
seconds for BB-BC. For a 100-node network, we found that minimum error achievable by BBO
after 100 iterations was 2.38% as compared to 1.36% of BB-BC. The time taken by BBO for 100
iterations was 9.55 seconds against 0.58 seconds in case of BB-BC. As the number of nodes is
increased to 2500, BB-BC took 21.26 seconds for 100 iterations to enumerate a shortest path with
1.90% error. BBO in contrast took 1 hour 12 minutes 29 seconds for the same number of
iterations and produced 2.982% error.
Throughput
End To End
Delay
Average Jitter
Fuzzy Inference
System
Cost of the Link

35
0 10 20 30 40 50 60 70 80 90 100
0.52
0.54
0.56
0.58
0.6
0.62
0.64
Generation
MinimumCost
Generation Vs Minimum Cost for 25 Nodes
BBO
BB-BC
Figure 5: Generation vs Minimum Cost for 25 Node network for 100 generations
0 10 20 30 40 50 60 70 80 90 100
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
Generation
MinimumCost
BBO
BB-BC
Figure 6: Generation vs Minimum Cost of 64 Node network for 100 generations

36
0 10 20 30 40 50 60 70 80 90 100
6
6.2
6.4
6.6
6.8
7
7.2
7.4
7.6
7.8
Generation
MinimumCost
BBO
BB-BC
Figure 7: Generation vs Minimum Cost of 2500 Node network for 100 generations
Table 1: Results of BB-BC and BBO algorithms
No. of
Nodes
No. Of
Generations
BB-BC BBO
Path
Cost
%
Error
Time (sec) Path
Cost
%
Error
Time (sec)
25
30 0.5350 0.00 0.104528 0.5350 0.00 1.813251
50 0.5350 0.00 0.139245 0.5350 0.00 2.194382
100 0.5350 0.00 0.214183 0.5350 0.00 2.584156
64
30 0.7315 1.147 0.224916 0.7442 2.903 2.761474
50 0.7232 0.00 0.271387 0.7242 0.138 3.420020
100 0.7232 0.00 0.289708 0.7232 0.00 4.960113
100
30 0.9197 6.05 0.325719 0.9274 6.94 3.977115
50 0.8990 3.66 0.407832 0.9135 5.33 6.76877
100 0.8790 1.36 0.581593 0.8879 2.38 9.557044
2500
30 6.3266 4.587 7.450026 6.7582 11.722 1278.180244
50 6.2640 3.55 11.174060 6.6384 9.741 1990.837452
100 6.1642 1.90 21.268608 6.2295 2.982 4249.000040

37
6. Conclusion and Future Scope of the Work
As the network complexity grows, time required to evaluate the exact shortest path increases.
This paper presented two recent soft computing based approaches to evaluate near shortest path in
a WMN. WMNs being highly dynamic need shortest path periodically to modify their routing
strategies. Further the shortest path evaluation technique must be able to compute a near shortest
path as quickly as possible so as to keep the routing performance optimal. Keeping the routing
policies in mind we proposed an integrated path length measure that takes into account
throughput, end-to-end delay and jitter of the link. A fuzzy based approach was used for
computing the distance between two adjacent nodes in terms of this integrated metric. Using this
integrated cost function as the path length metric we evaluated shortest path between source and
destination node using BB-BC and BBO approaches.
A large number of experiments were conducted to find the shortest path in a 25, 64, 100 and 2500
node models. It was observed that BB-BC approach clearly outperformed BBO approach in terms
of evaluation time and error between the actual shortest path and the evaluated near shortest path,
evaluated by these two approaches. For the small WMNs of 25, 64 and 100 nodes though the
performance of the two was comparable with BB-BC always producing better results. For a 2500
node model BB-BC produced its shortest path with 1.90% error in 21.26 seconds after 100
iterations whereas BBO generated its best path with 2.982% error in 1 hr 12 minute and 29
seconds. Thus, clearly establishing the superiority of BB-BC over BBO for finding the optimal
path in a WMN.
REFERENCES
[1] Akyildiz IF,Wang X and Wang W “Wireless mesh networks: a survey”, Computer Networks Journal
(Elsevier) 47(4), 2005, 445–487.
[2] C.Perkins, E.Belding-Royer and S.Das, “Ad hoc on-demand distance vector (AODV) routing”, IETF
RFC 3561, July 2003.
[3] Ogier R, Templin F and LewisM, “Topology dissemination based on reverse-path forwarding
(TBRPF)”, IETF RFC 3684, 2004
[4] D.B.Johnson, D.A.Maltz and Y.C.Hu, “The dynamic source routing protocol for mobile ad hoc
networks (DSR)”, IETF, July 2004.
[5] L. Huang, T. Lai, “On the scalability of IEEE 802.11 ad hoc networks”, ACM International
Symposium on Mobile Ad Hoc Networking and Computing (MOBIHOC), 2002, pp. 173–182.
[6] Akyildiz IF,Wang X and Wang W “Wireless mesh networks”, Wiley, 2009.
[7] A.Adya, P.Bahl, J.Padhye, A.Wolman and L.Zhou, “A multi radio unification protocol for IEEE
802.11 wireless networks”, International Conference on Broadcast Networks (Broad Nets), 2004,
pp.344- 354.
[8] Yang Y, Wang J and Kravets, “Interference-aware load balancing for multihop wireless networks”, In
Tech. Rep. UIUCDCS-R-2005-2526, Department of Computer Science, University of Illinois at
Urbana-Champaign, 2005.
[9] De Couto DSJ, Aguayo D, Bicket J and Morris R , “A high-throughput path metric for multihop
wireless routing”, In Proc. ACM Annual International Conference on Mobile Computing and
Networking (MOBICOM), 2003, pp. 134–146.
[10] Jakllari G, Eidenbenz S, Hengartner N, Krishnamurthy S and Faloutsos M, “Link positions matter: a
noncommutative routing metric for wireless mesh networks”, In Proc. IEEE Annual Conference on
Computer Communications (INFOCOM) 2008, pp. 744-752.
[11] R.Draves, J.Padhye and B.Zill, “Routing in Multi Radio, Multi-Hop Wireless Mesh Networks,” ACM
Annual International Conference on Mobile Computing and Networking (MobiCom’04), pp. 114-128.
[12] Karbaschi G and Fladenmuller A ,“A link quality and congestion-aware cross layer metric for multi-
hop wireless routing”, In Proc. of IEEE MASS’05, pp. 7–11.

38
[13] Ma L, Zhang Q, Xiong Y and Zhu W ,”Interference aware metric for dense multi-hop wireless
network”, In Proc. of IEEE International Conference on Communications (ICC), 2005, pp. 1261–1265.
[14] Koksal CE and Balakrishnan H, “Quality-aware routing metrics for time-varying wireless mesh
networks” IEEE Journal on Selected Areas in Communications 24(11), 2006 pp. 1984–1994.
[15] Liu T and Liao W, “Capacity-aware routing in multi-channel multi-rate wireless mesh networks”, In
Proc. IEEE International Conference on Communications (ICC), 2006, pp. 1971–1976.
[16] R. Draves, J. Padhye, B. Zill, “Comparisons of routing metrics for static multi-hop wireless networks”,
ACM Annual Conference of the Special Interest Group on Data Communication (SIGCOMM), August
2004, pp. 133–144.
[17] Shengxiang Yang, Member, IEEE, Hui Cheng, and Fang Wang, “Genetic Algorithms With Immigrants
and Memory Schemes for Dynamic Shortest Path Routing Problems in Mobile Ad Hoc Networks”,
IEEE Transactions on Systems, MAN, and Cybernetics—Part C: Applications and Reviews, vol. 40,
No. 1, January 2010, pp. 52-63.
[18] M. Kripka and R.M.L. Kripka, “Big Crunch Optimization method” International Conference on
Engineering Optimization, Brazil, June 2008,.
[19] O.K.Erol and I.Eksin, “A new optimization method: Big Bang-Big Crunch” Advances in Engineering
Software, 37, 106-111, 2006.
[20] T. Kumbasar, E. Yesil, I.Eksin and M. Guzelkaya, “Inverse fuzzy model control with online adaptation
via Big Bang-Big Crunch optimization,” ISCCSP 2008, Malta, 12-14 March, 2008, pp. 697-702.
[21] R. MacArthur and E. Wilson, “The Theory of Biogeography”, Princeton Univ. Press, 1967.
[22] D.Simon, “Biogeography-Based Optimization,”, IEEE Trans. on Evolutionary Computation, vol. 12,
no. 6, pp. 702-713, December 2008.
[23] Haiping Ma, “An analysis of the equilibrium of migration models for biogeography-based
optimization”, Information Sciences, 180 (2010), pp. 3444-3464, 2010.
[24] Anil Kumar, Arun Khosla, Jasbir Singh Saini and Satvir Singh “Meta-Heuristic Range Based Node
Localization Algorithm for Wireless Sensor Networks”, 2012 International Conference on Localization
and GNSS, pp. 1-7, Starnberg, Germany, June 25-27, 2012.
[25] Scalable Network Technologies, Inc. QualNet Simulator: http://www.scalable-networks.com.
[26] John Yen and Reza Langari, “Fuzzy Logic Intelligence, Control and Information,” Prentice Hall, New
Jersey, 1999.
Authors
Sharad Sharma, received his B.Tech in Electronics Engineering from Nagpur University,
Nagpur, India in 1998 and M.Tech in Electronics and Communication Engineering from
Thapar Institute of Engineering and Technology, Patiala, India in 2004. He has a teaching
experience of 14 years. He has conducted many workshops on Soft Computing and its
applications in engineering, Wireless Networks, Simulators etc. He has a keen interest in
teaching and implementing the latest techniques related to wireless and mobile
communications. He opened up a student chapter of IEEE as Branch Counselor. Presently, he is working
towards the Ph.D. degree in Electronics and Communication Engineering from National Institute of
Technology, Kurukshetra, India. His research interests are routing protocol design, performance evaluation
and optimization for wireless mesh networks using nature inspired computing.
Prof.(Dr.)Shakti Kumar received his M.S.(Electronics and Control) from Birla Institute of
Technology & Science (BITS), Pilani and Ph.D.(Electronics and Computer Engg.) from
National Institute of Technology (Formerly REC), Kurukshetra. He has served as a faculty
member in the department of Electronics & Computer Engineering, NIT Kurukshetra,
Department of Computer Engineering, Atlim University, Ankara, Turkey and Alghurair
University, Dubai, UAE. He was Co-ordinator, Centre for Excellence in Reliability
Engineering at REC (Now National Institute of Technology), Kurukshetra. He has published more than 165
research articles in various International/National journals and conferences, 7 book chapters and has
published one patent. He has chaired 13 International/National Conferences. He is member of various
professional bodies such as IEEE and ISTE. Dr. Kumar was jointly awarded “Sir Thomas Ward
Memorial” Subject Award by Institution of Engineers (India) in December 2006.

39
Dr. Brahmjit Singh, received his B.E. Degree in Electronics Engineering from Malaviya
National Institute of Technology (then MREC), Jaipur in 1988, M.E. with specialization in
Microwave and Radar from Indian Institute of Technology, Roorkee (then UoR) in 1995 and
Ph.D. from Guru Govind Singh Indraprastha University, Delhi in 2005. He is with
Electronics and Communication Engineering Deptt., National Institute of Technology
working as Professor since 2006. He has published 75 research papers in
International/National journals and conferences. His current research interests include mobility
management in mobile cellular networks, security in wireless networks, wireless sensor networks and
cognitive radio. He has been awarded “Sir Thomas Ward Memorial” Subject Award for the paper titled,
“Optimizing Handover Performance in Microcellular Systems” on behalf of Institution of Engineers (India)
in December, 2006. He is the member of IEEE, IETE and ISTE.

Routing in Wireless Mesh Networks: Two Soft Computing Based Approaches

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Routing in Wireless Mesh Networks: Two Soft Computing Based Approaches (20)

Recently uploaded (20)

Routing in Wireless Mesh Networks: Two Soft Computing Based Approaches