Wireless aggregated data then transmits to the sink either
Wireless sensor network (WSN) incorporates billions of sensing nodes that collaborate with each other to perform some sensing tasks and convey that data to the sink or Base Station (BS). Sensor nodes are deployed randomly in an ad-hoc manner 1. WSNs can be used in many applications such as battlefield surveillance, healthcare monitoring, landslide detection, forest fire detection, etc. The more accurate and reliable data can get by deploying a large number of sensor nodes. They have a limited constraint of battery, as they are deployed in a harsh environment, so it would be difficult to recharge or replace the sensor node batteries. So, the consumption of the node’s energy is to be properly administered so as to prolong the lifetime of the network. Since the lifetime of the network totally depends on the consumption of the energy of the node 2.
To improve the energy efficiency of WSNs, the typical method includes the energy efficient clustering and routing algorithms 3-4. Clustering is a technique used for periodic collection of data. In this technique, the network is divided into clusters with one leader is elected known as Cluster Head (CH) 5. All other nodes that are not elected as CH become cluster members (CM). Nodes sense the physical phenomena and send the data to the CH then the CH collects the data from all cluster members of its cluster and performs aggregation. The aggregated data then transmits to the sink either by single hop or multi-hop communication.
Clustering can be of two types namely, equal clustering and unequal clustering 6-7. In equal clustering, clusters of equal sizes are formed while in unequal clustering, clusters near the BS are less in size to that of clusters which is far away from BS.
Energy hole or hotspot issue is an issue in which sensor nodes near the BS dies prematurely. Since nodes near the sink sometimes act as a relay node, so dissipates more energy. Unequal clustering is used to cope up with this issue 7.
Designing an energy efficient clustering protocol is an NP-hard problem 8. To design this, various computational intelligence methods like fuzzy logic, neural networks, genetic algorithms, particle swarm optimization etc have been employed in WSNs for various issues 9-12. To solve the uncertainties in the WSNs, generally fuzzy logic is used. A system can be made more optimized with incomplete information with the use of fuzzy logic 13.
A fuzzy logic consists of three different parts 14. They are fuzzifier, inference engine, a fuzzy rule base and a defuzzifier. The input crisp value is converted into an appropriate fuzzy linguistic variable by fuzzifier block. Inference engine component processed the fuzzified values. A fuzzy rule component is simply a set of if-then rules that are needed to convert inputs into outputs. Defuzzifier block defuzzify the fuzzy linguistic output variable into crisp output value by using a suitable defuzzification method.
i. Main contribution and organization of the paper
An unequal clustering is proposed in this paper to prolong the lifetime of the sensor network. This scheme uses the fuzzy logic approach. Four inputs are given to the system, namely, residual energy, distance to base station, node degree and centrality. Fuzzy if-then rules are applied to these inputs to generate two outputs namely, competition radius and rank. The proposed protocol balances the load among the nodes in the networks and also prolongs the network lifetime by minimizing the dissipation of the energy. The main contributions of this paper are:
· It is a fuzzy based distributed clustering algorithm designed for the large-scale nonuniform network.
· It blends the probabilistic approach with the fuzzy logic approach in a suitable manner.
· Each node calculates the probability of being elected as a CH by the fuzzy logic approach.
· Clusters of variable sizes are also formed using fuzzy logic approach.
The remaining of the paper is as follows:
Section 2 describes the related work done so far in a comprehensive manner. Section 3 represents the network topology and the model used for experimental analysis. Section 4 gives the detailed description of the proposed protocol. Section 5 gives the analysis of the protocol. Section 6 depicts the simulation environment and the parameters used for simulation. Section 7 analyses the experimental results. Section 8 gives the conclusion followed by some future works.