Fuzzy Logic based Clustering Algorithm to Improve DEEC Protocol in Wireless Sensor Networks

Networking systems composed of wireless sensors that have limited power are commonly referred to as Wireless Sensor Networks. The primary aim of deployment of these sensing devices is to gather data from their environment and physical surroundings. While doing so, by virtue of their small and lightweight nature, they optimize the amount of energy dissipated, thereby differentiating them considerably from other typical networks. To fulfil this aim of transmission of data, lot of algorithms have been developed for carrying out routing. These protocols aim to enhance the overall network stability and lifetime. Clustering Based routing, a subset of hierarchical routing protocols, which include algorithms like the LEACH, SEP and DEEC is one such popular approach. In this approach, there is a single head node whose responsibility is data collection, i.e. gathering data from its entire cluster and then further sending it, passed through different aggregation functions, to the base station. In this paper, a new approach for clustering of the sensors has been proposed. It is an improvement of DEEC protocol. Fuzzy logic has been applied for the selection of the most suitable nodes to become the cluster head. A few more constraints have been considered for a particular node to be eligible to become the cluster head. The proposed approach outperforms traditional routing algorithms like SEP, LEACH and DEEC when compared with respect to overall network lifetime, stability period, network throughput and average energy.

I. Introduction

Wireless Sensor Networks (WSNs) [1] are networking systems comprising of compact sensor nodes that are required for the sensing of surrounding environmental conditions. These sensors are autonomous devices that are distributed spatially and monitor different parameters related to conditions of their physical surroundings or environment like pressure, temperature and humidity. In general, a sensor node is comprises of sensors, battery source of power, a micro-controller, memory and a transceiver. The architecture of the sensor node has been illustrated in figure 1. They are connected to each other in a wireless manner and also to a common node called the Sink (Base Station). In a WSN, each node may have the same energy or even different energies. The former case is called a homogenous network, while the latter a heterogeneous network. A principal feature of the working of these nodes is cooperation as all the different nodes collect and measure different kinds of mentioned parameters and transmit them to the Sink where they are further used in different applications. Today being one of the fastest evolving technologies, wireless sensor networks have found to be of great use in various avenues, both civil and military, like industrial monitoring, environment monitoring, surveillance etc. On one hand, these are tiny in size and wireless in nature and hence are easily deployed, even in difficult terrains, while on the other hand, their monitoring is difficult and power is very limited. Hence, to overcome these challenges of limited power and optimizing energy consumption, routing must be dealt with carefully.