You are on page 1of 5

HYBRID ROUTING PROTOCOL

OPTIMIZATION FOR ENHANCED LINK


QUALITY
Dr.V. Balaji1, N. Umapathi2, T. Karthikeyan 1, A.P. Prabakaran1
Faculties in Department of Electronics and Communication Engineering, A.V.C College of Engineering, Mayladuthurai
2
Faculty in Department of Electronics and Communication Engineering, G.K.M College of Engineering, Chennai
1
balajivece@avccengg.net, 2nrumapathi@gmail.com

Abstract The combinatorial optimization problem is

applied to various engineering fields. There is the shortest path


problem in one of the combinatorial optimization problems. It
can be classified into two categories. One is the problem for
the searching on the grid and the other is that for the searching
on the graph. In this paper, a new routing algorithm using
ACO is proposed. The proposed algorithm deals with the route
searching problem considering link quality constraints. There
is a possibility that the ant agent is trapped in the blind alley.
Thus, the proposed algorithm combines Tabu search algorithm
with ACO to overcome this problem. The next section gives
brief introduction to Tabu Search.
Key words: ACO, Tabu search, link quality, Hybrid routing
Algorithms

I.

INTRODUCTION

The combinatorial optimization problem is applied to


various engineering fields. There is the shortest path problem
in one of the combinatorial optimization problems. It can be
classified into two categories. One is the problem for the
searching on the grid as shown in Fig.1 (1). The other is that
for the searching on the graph as shown in Fig.1 (2). Maze
algorithm (1) which was proposed by Lee is the routing
algorithm for the former category, and it is applied to VLSI
CAD problems. On the other hand, Dijkstra algorithm (2) is
the routing algorithm for the latter category, and it is applied
to route guidance problem such as car navigation system.

Figure 1: Representation of shortest path

These routing algorithms find the shortest path, whenever the


path exists. However, they have the inherent processing time
on the searching process. Recently, many researchers studied
the meta-heuristics which can find the sub-optimal solution at
short time. In these meta-heuristics, it is reported that Ant
Colony Optimization (ACO), which is inspired by feeding
behaviour of ants, shows the better capability than Genetic
Algorithm (GA) and Simulated Annealing (SA) when it is
applied to Travelling Salesman Problem (TSP).
II.

TABU SEARCH

Tabu search is based on introducing flexible memory structures


in conjunction with strategic restrictions and aspiration levels
as a means for exploiting search spaces. Meta-heuristic that
guides the local heuristic search procedure to explore the
solution space beyond local optimum by use of a Tabu list
(Becerra, et al., 2006).
Tabu search (TS) is dramatically changing the ability to
solve problems of practical significance. A distinguishing
feature of Tabu search is embodied in its exploitation of
adaptive forms of memory, which equips it to penetrate
complexities that often confound alternative approaches. More
particularly, Tabu search is based on the premise that problem
solving, in order to qualify as intelligent, must incorporate
adaptive memory and responsive exploration. The adaptive
memory feature of TS allows the implementation of
procedures that are capable of searching the solution space
economically and effectively. Responsive exploration
integrates the basic principles of intelligent search, i.e.,
exploiting good solution features while exploring new
promising regions. Tabu search is concerned with finding new
and more effective ways of taking advantage of the
mechanisms associated with both adaptive memory and
responsive exploration. The development of new designs and
strategic mixes makes TS a fertile area for research and
empirical study (Glover, et al., 1993).
Two highly important components of Tabu search are
intensification and diversification strategies. Intensification
strategies are based on modifying choice rules to encourage
move combinations and solution features historically found
good. They may also initiate a return to attractive regions to
search them more thoroughly. The diversification stage on the
other hand encourages the search process to examine unvisited

regions and to generate solutions that differ in various


significant ways from those seen before.
Principal Features of Tabu Search (Glover 1995) are:
Adaptive Memory
Selectivity (including strategic forgetting), Abstraction
and decomposition (through explicit and attributive memory)
Timing:
Recency of events, Frequency of events, Differentiation
between short term and long term, Quality and impact,
Relative attractiveness of alternative choices, Magnitude of
changes in structure or constraining, Relationships
Context:
Regional interdependence, Structural interdependence
,Sequential interdependence ,
Responsive Exploration
Strategically imposed restraints and inducements (Tabu
conditions and aspiration levels)
Concentrated focus on good regions and good solution
features (Intensification processes)
Characterizing and exploring promising new regions
(Diversification processes)
Non-monotonic search patterns (Strategic oscillation)
Integrating and extending solutions (Path re-linking)
2.1 TABU SEARCH TECHNIQUES
In rough terms the Tabu search method can be sketched
as follows: to move step by step from an initial feasible
solution of a combinatorial optimization problem towards a
solution giving the minimum value of some objective
function. For this each solution is represented by a point in
some space and has to be defined in a neighborhood N (s) of
each point s.
The basic step of the procedure consists in starting from
a feasible point s and generating a sample (with fixed size rep)
of solutions in N (s); then choose the best neighbor s*
generated so far and move to s* whether f(s*) is better than
f(s) or not.
Up to this point this is close to a local improvement
technique except the fact that it may move to a worse solution
s* from s (this is a situation which occurs in simulated
annealing where a move to a worse solution may be accepted
with a probability which decreases when the number of
completed iterations increases). The interesting feature of
Tabu search is precisely the construction of a list T of Tabu
moves: these are moves which are not allowed at the present
iteration. The reason for this list is to exclude moves which
would bring it back to where it was at some previous iteration.
Now a move remains a Tabu move only during a certain
number of iterations, so that in a cyclical list T where at each
move ss* the opposite move s* s is added at the end of T
while the oldest move in T is removed from T.
In conclusion the basic step consists in generating
randomly a fixed number rep of possible moves from s
(whenever a move in T is generated, it is destroyed and a new

move is generated). Then the best one of the generated moves


is realized and the Tabu list T is updated accordingly. Now a
stopping rule should be also defined: in general a maximum
number of iterations may be given (Hertz, et al., 1987).

Figure 2: Dimensions of Memory structure in Tabu Search


The memory structures in Tabu search operate by
reference to four principal dimensions, consisting of recency,
frequency, quality, and influence. Recency-based and
frequency-based based memory complements each other. The
quality dimension refers to the ability to differentiate the merit
of solutions visited during the search. In this context, memory
can be used to identify elements that are common to good
solutions or to paths that lead to such solutions. The fourth
dimension, influence, considers the impact of the choices
made during the search, not only on quality but also on
structure. (In a sense, quality may be regarded as a special
form of influence.) Recording information about the influence
of choices on particular solution elements incorporates an
additional level of learning.
The memory used in Tabu search is both explicit and
attributive. Explicit memory records complete solutions,
typically consisting of elite solutions visited during the search.
An extension of this memory records highly attractive but
unexplored neighbors of elite solutions. TS use attributive
memory for guiding purposes. This type of memory records
information about solution attributes that change in moving
from one solution to another (Glover, et al., 1993). The Flow
Chart is given below for the Tabu Search

Figure 3: Flowchart of Tabu Search

2.2 BASIC TABU SEARCH ALGORITHM


Step 1: Choose an initial solution i in S. Set i* = i and k=0.
Step 2: Set k=k+1 and generate a subset V* of solution in
N(i,k) such that either one of the Tabu conditions is violated or
at least one of the aspiration conditions holds.
Step 3: Choose a best j in V* and set i=j.
Step 4: If f(i) < f(i*) then set i* = i.
Step 5: Update Tabu and aspiration conditions.
Step 6: If a stopping condition is met then stop. Else go to
Step 2.
Some immediate stopping conditions could be the following:
N (i, K+1) = 0. (no feasible solution in the neighborhood of
solution i)
K is larger than the maximum number of iterations allowed.
The number of iterations since the last improvement of i* is
larger than a specified number.
Evidence can be given than an optimum solution has been
obtained.
Hillier and Lieberman outlined the Tabu search stopping
criterion by, for example, using a fixed number of iterations, a
fixed amount of CPU time, or a fixed number of consecutive
iterations without an improvement in the best objective
function value. Also stop at any iteration where there are no
feasible moves into the local neighborhood of the current trial
solution.
3 METHODOLOGY
The searching on ACO utilizes two evaluations which
consist of the static value and the dynamic one. The static
evaluation is peculiar information of the target problem.
Usually, a reciprocal number of the distance is adopted as the
static evaluation value, when ACO is applied to the route
searching problem. On the other hand, the dynamic evaluation
introduces pheromone amount.
Specifically, the optimization procedure of ACO is
explained using an example of which ACO is applied. First,
the random number q between from 0 to 1 is generated. Next,
q is compared with benchmark (parameter) q0. When q is
smaller than q0, the node that has the largest value of the
product of the static evaluation and the dynamic one is
selected as the next destination.
Otherwise, ant k in node i selects the move to node j
according to probability pk and it is defined as follows.

p k i, j

i, j

l , j

lnk

i, j

l , j

where, (i,j) is a pheromone amount between node i and node


j, (i,j) is a reciprocal of the distance between node i and node
j, is a parameter which controls the balance between static
evaluation value and dynamic one, and nk is a set of un-visited
nodes.
Regarding the dynamic evaluation, a pheromone amount
on each route between cities is calculated by using two
pheromone update rules. One is local update rule and the other

is global update rule. The local update rule is applied to the


route which is selected, and it is defined as follows.

i, j 1 i, j 0

where, is a decay parameter in local update rule, 0 is the


initial value of pheromone. Thus, the local update rule adds
the pheromone to the selected route, when the ant moves. The
global update rule adds pheromone to the best route (the
completed route) of all routes. The best route usually indicates
the shortest route. The global update rule is defined as follows.

i, j 1 i, j i, j
1 / L
i, j
0

if i, j T
otherwise

where, is s a decay parameter in the global update rule, T+


is the best route, and L+ is the distance of the best route.
When ACO is applied to the route searching problem, there
are the following two problems: (1) the moving destination
has been limited, and (2) the ant argent is trapped in the blind
alley. Regarding of the problem (1), the proposed algorithm
solves it by separately handling the un-visit node, the visited
node and the moving candidate node, respectively. In addition,
regarding the blind alley of the problem (2), the proposed
algorithm overcomes it by adopting Tabu search. Here, the
situation of which the ant agent is trapped in the blind alley
represents that the ant agent has not reached the final
destination and no moving candidate node exists. In other
words, the current node of which the ant agent stays has only
the connections to the visited nodes. However, the node d has
no moving candidate node as well as the previous situation. In
this case, the node d has only the visited node and the taboo
node. Therefore, the ant agent returns to the previous node as
well as the previous processing regarding the taboo operation.
These processing are repeated until the moving
destination candidate's node appears. The concrete procedure
is as follows.
Step1: Select the start node and the destination node (target
node).
Step2: Select the node using equations stated above.
Step3: Move to the selected node and mark the current node as
the visited node.
Step4: Taboo operation is executed when no moving candidate
node exists.
Step5: Repeat from Step2 to Step4 until the ant agent reaches
to the final destination.
Step6: Update the pheromone value.
Step7: Repeat from Step2 to Step6 until the generations are
terminated.
Thus, the proposal algorithm enables ACO to be applied to the
route searching problem by combining Tabu search with ACO.

4. RESULTS AND DISCUSSION

The experimental setup is similar to the setup used in the


previous chapter. The results present the performance of the
protocol in the following tables and graphs. The proposed
hybrid ACO-AODV is compared with AODV and ACOAODV. Figure 4 shows the Packet Delivery Ratio of the
proposed network using hybrid ACO and Figure 5 shows the
End to End delay. Table1 tabulates the PDR for the proposed
protocol.
Table 1: Packet Delivery Ratio for varying Node Pause Time
for the proposed Hybrid ACO-AODV

Table 2 shows the result of PDR and it is observed that the


PDR increases with the increase in pause time for both ACOAODV and Hybrid ACO-AODV. The performance of Hybrid
ACO-AODV improves by 5.22% when compared to AODV
and by 3.87% when compared to LQ-AODV when node pause
time is 20 sec. Similarly the performance of Hybrid ACOAODV improves by 5.63% when compared to AODV and by
3.34% when compared to LQ-AODV when node pause time is
100 sec. Table 2 tabulates the End to End Delay.
Table 2: End to End Delay for varying Node Pause Time for
the proposed hybrid ACO-AODV

Table 3: Jitter for varying Node Pause Time for the proposed
hybrid ACO-AODV

Table 3 gives the result of Jitter for pause times varying from
20 to 100 sec. It is observed that Jitter decreases with the
increase in pause time for all variations of AODV simulated.
Jitter of Hybrid ACO-AODV is less by 72.25% when
compared to AODV and by 38.15% when compared to LQAODV when node pause time is 20 sec. Similarly Jitter of
Hybrid ACO-AODV is less by 15.37% when compared to
AODV and by 1.66% when compared to LQ-AODV when
node pause time is 100 sec. Hybrid ACO-AODV has least
Jitter.
It is apparent from the above graphs and tables of the
simulation results of PDR, End to end delay and jitter that the
performance of the network improves with the proposed
enhancement of the AODV protocol. The hybrid optimization
of the routing has improved the performance of the network
with increase in PDR and decrease in end to end delay and
Jitter.
The same setup was used with Node pause time of 100
for finding route discovery time, total route cache sent and
total packet dropped for AODV, VO-AODV and LQ-AODV.
The simulations were run for 600 sec. Following Tables and
Figures show the simulation results.
Simulation
time (sec)

Table 2 gives the result of end to end delay for pause


times varying from 20 to 100 sec. It is observed that the end to
end delay decreases with the increase in pause time for all
variations of AODV simulated. The end to end delay of
Hybrid ACO-AODV is less by 8.86% when compared to
AODV and by 3.06% when compared to LQ-AODV when
node pause time is 20 sec. Similarly the end to end delay of
Hybrid ACO-AODV is less by 24.16% when compared to
AODV and by 17.56% when compared to LQ-AODV when
node pause time is 100 sec. Hybrid -AODV has least end to
end delay. Hybrid ACO-AODV has least end to end delay
decreases. Table 3 tabulates the Jitter for the network.

Table 4: Route Discovery Time


LQ-AODV
ACO-AODV
HYBRIDAODV

18

0.053319313

0.055337633

0.01223088

24

0.005723108

0.007965677

0.017806363

102

0.009980489

0.007185544

0.007041833

360

0.005848126

0.005052016

0.004950976

408

0.006397806

0.007319559

0.007173168

486

0.005469234

0.041274077

0.00404486

552

0.004375336

0.037221393

0.006476965

582

0.007015686

0.004485955

0.004396236

594

0.00362204

0.005549013

0.005338033

Table5: Total Route Cache Sent

simulation
time(sec)
0
18
24
102
114
186
282
288
294
300
simulation
time(sec)
0
18
24
120
174
228
252
270
294
300

LQAODV
0
12
16
10
26
12
35
26
24
23

ACOAODV
0
3
6
206
178
351
205
138
145
189

HYBRIDAODV
0
5
7
187
180
290
202
148
149
179

Table 6: Total Packet Dropped


LQACOHYBRIDAODV
AODV
AODV
0
0
0
5
5
3
12
17
11
16
20
14
24
9
5
18
16
17
24
19
20
23
22
18
12
10
9
0
10
9

5 Conclusion
Major swarm intelligence research focused on reverse
engineering and collective behaviors adaptation observed in
natural systems aimed at effective algorithm design for
distributed optimization. Such algorithms inspired by natural
systems reveal desirable properties like adaptability,
scalability and robustness which are key properties in network
routing, specifically wireless network routing. Designers
conventionally depended on optimization methods that neither
need prior knowledge nor centrally managed runtime
knowledge about the systems environment, since these
methods are feasible in dynamic networks as there is often
unpredictable modifications in topology, resource, and node
availability. In this study, a hybrid optimization for routing
based on ACO and Tabu Search was proposed. Simulations
were run and the proposed routing was compared with LQAODV and ACO-AODV. An improvement in packet delivery
ratio and end to end delay is observed in the proposed Hybrid
ACO- AODV.
REFERENCES
[1] Glover, F., & Laguna, M. (1993). Tabu search (pp. 70150). John Wiley & Sons, Inc.
[2] Hertz, A., & de Werra, D. (1987). Using tabu search
techniques for graph coloring. Computing, 39(4), 345351.
[3] Becerra, J. G. G., Amado, R. J. A., & y Administracin, E.
D. I. A TABU SEARCH APPROACH FOR THE
TRAVELING SALESMAN PROBLEM.

[4] Glover, F. (1995). Tabu search fundamentals and uses.


Boulder: Graduate School of Business, University of
Colorado.
[5] Kim, S., Lee, O., Choi, S., & Lee, S. J. (2011).
Comparative analysis of link quality metrics and routing
protocols for optimal route construction in wireless mesh
networks. Ad Hoc Networks, 9(7), 1343-1358.
[6] Ikeda, M., Barolli, L., Hiyama, M., De Marco, G., Yang,
T., & Durresi, A. (2009, March). Performance evaluation
of link quality extension in multihop wireless mobile adhoc networks. In Complex, Intelligent and Software
Intensive Systems, 2009. CISIS'09. International
Conference on (pp. 311-318). IEEE.
[7] Fathima, K. S. A., Sindhanaiselvan. K. (2013). Ant
Colony Optimization Based Routing in Wireless Sensor
Networks. Int. J. Advanced Networking and
Applications, 4(04), 1686-1689.
[8] Biradar, R., Manvi, S., & Reddy, M. (2010). Link stability
based multicast routing scheme in MANET. Computer
Networks, 54(7), 1183-1196.
[9] Mamun-Or-Rashid, M., & Hong, C. S. (2007, May).
LSLP: link stability and lifetime prediction based qos
aware routing for MANET. In Joint Conference on
Communications and Information, JCCI.
[10] Clausen, T., Jacquet, P., & Dearlove, C. (2012). Link
Metrics for the Mobile Ad Hoc Network (MANET)
Routing Protocol OLSRv2-Rationale.
[11] Oh, S. Y., Gerla, M., & Tiwari, A. (2009, January). Robust
MANET routing using adaptive path redundancy and
coding. In Communication Systems and Networks and
Workshops, 2009. COMSNETS 2009. First International
(pp. 1-10). IEEE.
[12] Farkas, K., Hossmann, T., Ruf, L., & Plattner, B. (2006,
October). Pattern matching based link quality prediction
in wireless mobile ad hoc networks. In Proceedings of
the 9th ACM international symposium on Modeling
analysis and simulation of wireless and mobile systems
(pp. 239-246). ACM.
[13] Nakaoka, K., Oba, M., & Tasaka, S. (2006, September).
LQHR: A link quality-based hybrid routing protocol for
wireless ad hoc networks. In Personal, Indoor and
Mobile Radio Communications, 2006 IEEE 17th
International Symposium on (pp. 1-8). IEEE.
[14] Ali, H. M., Naimi, A. M., Busson, A., & Vque, V. (2007,
October). An efficient link management algorithm for
high mobility mesh networks. In Proceedings of the 5th
ACM international workshop on Mobility management
and wireless access (pp. 42-49). ACM.
[15] Zhou, J., Jacobsson, M., Onur, E., & Niemegeers, I.
(2012). An investigation of link quality assessment for
mobile
multi-hop
and
multi-rate
wireless
networks.Wireless personal communications, 65(2), 405423.

You might also like