International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]

Volume 4, Issue 1, January-February 2015
ISSN 2278-6856

Collating Techniques of Computational

Intelligence Paradigms: An Ingenious
Approach
Nidhi Bhartiya1, Dr. S.Taruna2
1

Department of Information Technology, Banasthali Vidyapith,

Jaipur, Rajasthan, India
2

Department of Computer Science, Banasthali Vidyapith,

Jaipur, Rajasthan, India

Abstract
Wireless Sensor Network (WSN) is a collection of thousands
of spatially disperse nodes that can sense various physical or
environmental characteristics such as light, temperature etc.
WSNs face many issues, especially due to failures in
communication network, storage of data and limited power
supply. Multiple Paradigms of computational intelligence (CI)
had been addressed various challenges successfully in recent
years and by using these paradigms many limitations will also
be reduced. In this paper some main challenges and
comparison between different paradigms on the basis of
several issues are defined. In this paper some advantages and
disadvantages are also defined.

Keywords: WSN, Fuzzy logic, Neural Network, AIS,

Swarm Intelligence.

1. INTRODUCTION
Wireless sensor network is a collection of thousands
number of sensor nodes. These sensor nodes sense the
data on the basis of various parameters like temperature,
distance etc. Typically, sensor nodes are collected in a
form of group which is known as cluster and each cluster
has a node that acts as the cluster head. All nodes transmit
their sensory data to the cluster head, which in turn routes
it to a specialized node (or base station) through a multihop wireless communication. When sensor nodes transmit
the data between nodes to base station, various challenges
are defined. Researchers have used these CI techniques to
find out the solutions and they will also address many
issues in WSN.

2. CHALLENGES IN SENSOR NETWORKS

For deployment of nodes in WSN generally used one of
the given applications[1]:
Periodic reporting
Event detection
Database storage.
Periodic reporting is one of the simplest application in
which data can be sensed and stored at regular intervals
and send it to the base station(s). This is used in most
monitoring applications in agriculture, military
operations, and disaster relief. Prediction about traffic in

Volume 4, Issue 1, January February 2015

data is the most important feature of periodic reporting. In

event detection applications, nodes sense the environment
and find out that data is useful or not. If data is found
useful, it will transmit to the base station(s). In third, all
data which was sensed according to environment is stored
locally on the nodes. When Base stations need that data
they can search and retrieve it directly from the nodes.
The main issue in these applications is to storing the data
so when related data is search it will give faster and better
outcome. Some major WSN issues addressed by CI
techniques are discussed in the following subsections:
A. Design and deployment
In WSN, to get a complete information network are
designed according to the requirement So for getting
better results sensor node should be placed accurately at
their proper places.
B. Localization
Node localization is used to detect the location of sensor
nodes. Only awareness of location is not sufficient beside
that data is also need to be sensed by nodes. To search and
store any event or record or any information about
location of nodes geometric routing is used.
C. Data Aggregation and Sensor Fusion
Collection of data from the various sources at a single
place definitely it will provide some better result but also
increase the communication overhead. Due to nodes
deploy at a very large scale, data is also generated in a
huge volume, so collecting data in an efficient manner is a
big challenge. Most widely used non-CI methods for
sensor fusion include Dempster-Shafer method, Bayesian
networks.
D. Energy Aware Routing and Clustering
Energy is one of the most important issues in WSN. In
some areas lifetime of battery is playing an important role.
Changing the battery very frequently is very dangerous
and it will also increase the cost. Routing is defined in
such a way that it determined a proper route for
transmitting a message from source to its particular
destination. Routing is classified in two categories, first
one was proactive and second one was reactive routing. In
proactive routing methods, whenever the routes are used,
it will store in the form of table. In reactive routing
Page 142

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 1, January-February 2015
ISSN 2278-6856
methods, routes are computed as according to its
requirement.
E. Security
Security is also one of the most important issues in WSN.
Main key issue of the security was to secure or protect the
data from any type of internal or external attacks and to
provide an adequate security. Administration becomes
more difficult and applies various security techniques or
methods to provide more security so data will be more
secure. Now days, various security challenges in WSN are
addressed and to protect our data from these intruders is a
key issues.

3. PARADIGMS OF CI
Computational Intelligence contains various paradigms
such as neural networks, SI, evolutionary algorithms,
fuzzy logic and AIS. These paradigms are described in the
following subsections
3.1 Neural Networks
Human brain is an excellent ability to learn and memorize
any new thing very easily and in a faster manner.
Artificial neural network is defined such as, it is a parallel
computing systems, consists of large number of
interconnected neurons having simple processors to solve
a computational challenges with the help of human brain.
Neural network works on three basic components are as
follow [6]:
1) Wji provide weight to the n inputs of jth neuron xi,
i = 1,2, . . . , n;
2) To compute the sum of input of activation function an
aggregation function is used, uj =, where j is the bias,
which is a numerical value associated with the neuron.

Fig 2: Fuzzy Structure

3.3 Evolutionary Algorithms

Evolutionary algorithms are stochastic search methods
which defines the natural biological evolution. EAs use a
population of solution candidates called chromosomes.
Chromosomes are a combination of genes, which
represent a different characteristic. At each level next
level chromosomes will be choose on the basis of fitness.
In reproduction process two parent chromosomes are mix
into a new one which is called as offspring. Chromosomes
which having highest fitness is selected and it will go into
the next generation, and the rest are discarded. The
process is repeated until a final result cant be obtained.

3) An activation function that maps uj to vj = (uj),the

output value of the neuron

Fig 3: Structure of evolutionary algorithm

3.4 Swarm Intelligence
Swarm Intelligence (SI) is one of the most powerful
paradigms of computational intelligence [4,10]. Swarm
intelligence is defined as a study of behavior of biological
species such as colonies of ants. SI is also very compatible
with WSN routing. Due to environment ontological
match, various efficient routing techniques can be
addressed by the help of SI and WSN.

Fig 1: Structure of Neuron

3.2 Fuzzy Logic
Fuzzy deals with uncertainty or imprecise data, which is
represented by the use of linguistic variables such as most,
few, very few etc. Fuzzy logic is a framework in which it
includes fuzzy sets, fuzzy rules and membership functions.
Fuzzy rules are defined in if-then format[11]. For
example-if service is good then food is delicious.
Knowledge base is formed with the combination of fuzzy
rules and fuzzy sets

Volume 4, Issue 1, January February 2015

In SI, a collective behavior of agents is collected in a

system which interacts with environment locally and
cause coherent functional global pattern to emerge. Highly
dynamic behavior of biological species is addressed in
decentralized and self-deployed systems.
3.4.1 Particle Swarm Optimization
PSO consists of a swarm of s particles, which represent a
best candidate solution among them. Each particle works
on two factors, first was velocity id and second one was
position id. Swarm particles are randomly assigned an
initial level and then calculate the fitness for each
particle[9]. PSO system combines local search methods
with global search methods, attempting to balance
Page 143

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 1, January-February 2015
ISSN 2278-6856
exploration and exploitation As such a particle is near to
the solution its fitness value is as much higher compared
to other particles. The main objective of Particle Swarm
Optimization is to find the location of an optimal point in
a Cartesian coordinate system. PSO work as follows:

Fig 4: Flow chart of PSO

3.4.2 Ant Colony Optimization

find out best path among all of them. They communicate

each other by the help of pheromone and they choose their
path with respect to previously available probabilities.[12]

3.5 AIS
AIS are abbreviation of Artificial Immune System. This is
a paradigm of computational intelligence which is used
for solving a problem. This approach is inspired from the
biological human capabilities to perform some
computation task. Recently AIS shows that clustering and
cluster head selection can be efficiently applied. Main
problem of using routing and clustering in WSN are fixed
number of clusters. AIS are a very new area in field of
research so it cant be well studied yet. In Paper, mainly
concentrate on the number of cluster heads and to
minimizing the energy consumption of the overall
transmission. In this, researcher minimizes the
communication distance with guarantee that it will
increase the lifetime of network. It concentrate mainly on
reduce the level of used energy in each node. In this main
priority was to reduce the level of consumed energy and
extend the lifetime of network.
Table 1: Comparison on the basis of energy conservation

ACO was defined as a metaheuristic approach, which is

used for finding the solution of the optimization problem.
The basic idea of the ant colony optimization (ACO) is
searching the food by using the shortest path without
directly communication. Real ants are communicated
indirectly with the help of pheromone. Ants are classified
in two categories: First one was FANT and second one
was BANT. In ACO approach ants find out the good
quality of food in a good quantity and stored it to their
nest when they return to their nest they use trail of
chemical pheromone, which also guides other ants to
reach that particular place where the food source is
stored[8]. ACO works in a following manner:

Fig 5: Flow chart of ACO

The main idea of the ACO metaheuristic is to find out the
best path in a construction graph" that represents the
states of the problem. Artificial ants used this graph and

Volume 4, Issue 1, January February 2015

Page 144

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 1, January-February 2015
ISSN 2278-6856
Table 2:Comparative Analysis of CI Algorithm Evolution
of Energy Aware Routing in WSN

4. Conclusion
In this paper various paradigms of CI are discussed along
with their issues and challenges. CI paradigms propose
non-conventional approach to solve the various issues in
different fields and in WSN. Comparative analysis of CI
paradigm algorithms shows algorithms which can be best
suited for energy aware routing and optimization criteria
in WSN can be a hybrid techniques which can be the
solution for optimized energy aware routing techniques to
make computationally efficient and better choice for WSN
optimized energy aware routing

References

Table 3: Advantages and Disadvantages of CI Paradigms

Volume 4, Issue 1, January February 2015

[1] R.V. Kulkarni, A. Frster, G.K. Venayagamoorthy,

Computational Intelligence In
Wireless Sensor
Networks: A Survey, IEEE Communications Surveys
& Tutorials, Vol. 13, No. 1, First Quarter 2011
[2] R. Iram, M.I. Sheikh, S. Jabbar, A. Minhas,
Computational Intelligence Based Optimization Of
Energy Aware Routing In WSN, Proceedings Of The
World Congress on Engineering and Computer
Science, WCECS 2011, October 2011.
[3] Sujatha .S.R. Dr.M.Siddappa, A Study on Efficient
Localization Methods Using Swarm Intelligence,
International Journal of Research in Computer
Applications and Robotics, Vol 2, 2014.
[4] M.Sharawi, I.A. Saroit, H. Mahdy, E. Emary,
Routing Wireless Sensor Networks Based On Soft
Computing Paradigms: Survey, International Journal
on Soft Computing, Artificial Intelligence and
Applications (IJSCAI), Vol.2, No.4, August 2013.
[5] Sachin Gajjar, Mohanchur Sarkar, Kankar Dasgupta,
Cluster Head Selection Protocol Using Fuzzy Logic
For Wireless Sensor Networks, International Journal
Of Computer Applications (0975 8887) Volume 97
No.7, July 2014.
[6] http://dspace.cusat.ac.in/jspui/bitstream/123456789/3
184/1/Computational_Intelligence_Seminar_Sruthy.p
df
[7] B. Solaiman, A. Sheta, Computational Intelligence
for Wireless Sensor Networks: Applications and
Clustering Algorithms, International Journal of
Computer Applications (0975 - 8887) Volume 73 No. 15, July 2013.
[8] S.U. Parvatkar and D.V. Gore, An Ant Colony
Optimization Algorithm for maximizing the lifetime
of Heterogeneous Wireless Sensor Network,
International Journal of Current Engineering and
Technology Vol.4, No.3 (June 2014).
[9] I. Maleki, S.R Khaze, M.M.Tabrizi and
A.Bagherinia, A New Approach For Area Coverage
Problem In Wireless Sensor Networks With Hybrid
Particle Swarm Optimization And Differential
Evolution Algorithms, International Journal Of
Mobile Network Communications & Telematics (
IJMNCT) Vol. 3 No.6, December 2013.
[10] Z. Ali, W. Shahzad, Analysis of Routing Protocols
in Ad Hoc and Sensor Wireless Networks Based on
Page 145

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 1, January-February 2015
ISSN 2278-6856
Swarm Intelligence, International Journal of
Networks and Communications 2013.
[11] Tarique Haider and Mariam, A Fuzzy Approach to
Energy Optimized Routing for WSNs, International
Arab Journal of Information Technology, Vol.6, No.2,
April 2009
[12] X. Wei and L. Zhi, The multi-objective routing
optimization of WSNs based on an improved ant
colony algorithm, in Proceedings of the 6th
International
Conference
on
Wireless
Communications, Networking and Mobile Computing
(WiCOM), September 2010.

Volume 4, Issue 1, January February 2015

Page 146