Biologically Inspired Clustering Mechanism in Dense Distributed Wireless Sensor Networks

A wireless sensor is a miniature component which measure physical parameters from the environment and transmit them to the monitoring station by wireless medium. In wireless medium, the sensor and its associated components are called as node. A node is self-possessed by a sensor, processor, local memory, transceiver and a low-powered battery. Wireless sensor networks are group of sensor nodes with a set of processors and limited memory unit embedded in it. Reliable routing of packets from sensor nodes to its base station is the most important task for these networks. Clustering is an important task for attaining some valued outputs like improved energy efficiency, reduced delay, increased throughput and reduced data losses. In order to produce well balanced clusters, the cluster head is rotated periodically with the help of a distributed algorithm. This paper gives a detailed study of various distributed clustering approaches. A detailed research is made on optimized cluster initialization based on jumping ant approach in order to avoid random cluster initialization. Also this mechanism shows directions on how to rotate the cluster head periodically and energy efficiently. The algorithm consists of three stages. In the first stage, the ants move towards the available food. In the second stage, the ants that gets sufficient food stays in that cluster. In the third stage, the foodless ants jumps and form another cluster. This mechanism clearly shows an excellent improvement over those with random initializations