A BAYESIAN APPROACH TO SELECT CLUSTER HEADS IN WIRELESS SENSOR NETWORKS
SEYED NASER RAZAVI *
Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran.
AIDA VALINEZHAD ORANG
Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran.
NAZLI BAGHERZADEH KARIMI
Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran.
*Author to whom correspondence should be addressed.
Abstract
Wireless sensor networks are composed of one or more Base Stations and many nodes. Each node of this network has a battery with limited energy. Since these nodes are used in places where accessing is difficult or impossible, recharging the battery is costly, or in some cases, it is practically impossible. Therefore, one of controversial and challenging subjects is consuming the energy of each node and consequently minimizing energy consumption in the network. In this article, an algorithm called Distributed Clustering based on Bayesian Networks (DCBN) has been proposed for nodes clustering by using Bayesian networks. The purpose of the proposed algorithm is to select the suitable Cluster Head in each cluster to find the optimal route to send data. At last, through reducing energy consumption, it has been attempted to increase the network lifetime. According to simulation results, the number of live nodes in DCBN algorithm is more than the three other algorithms LEACH, CROSS and DCGT. According to simulation results, death of nodes occur with delay in DCBN algorithm; as a result, the network lifespan increases in comparison to other three algorithms.
Keywords: Wireless sensor networks, clustering, Bayesian networks