Highly Scalable Energy Efficient Clustering Methodology for Sensor Networks
(IJAER), Vol. No. 12, Issue No. IV, October 2016
Posted: 1 Dec 2016
Date Written: November 25, 2016
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. To diminish the data transmission time and energy consumption, the sensor nodes are assembled into a number of little groups referred as clusters and the phenomenon is referred as clustering. Every cluster comprise of a leader which is known as cluster head. The cluster head will be chosen by the sensor nodes in the individual cluster or be pre-assigned by the user. The main advantages of clustering are the transmission of aggregated data to the base station, offers scalability for huge number of nodes and trims down energy consumption. Fundamentally, clustering could be classified into centralized clustering, distributed clustering and hybrid clustering. In centralized clustering, the cluster head is fixed. The rest of the nodes in the cluster act as member nodes. In distributed clustering, the cluster head is not fixed. The cluster head keeps on shifting form node to node within the cluster on the basis of some parameters. Hybrid clustering is the combination of both centralized clustering and distributed clustering mechanisms. 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.
Keywords: Wireless Sensor Network, Distributed Clustering Algorithm, Ant Colony Optimization, Jumping Ant, Energy Efficiency, Network Lifetime
Suggested Citation: Suggested Citation