Energy efficient and balanced clustering approach for improving throughput of wireless sensor networks

Abstract

Maximizing the lifespan of wireless sensor networks is presently drawing much attention in the research community. To reduce energy consumption, sensor nodes which is far away from the base station avoid sending data directly. As a result, several disjoint clusters are formed and nodes within the cluster send its data through cluster head to avoid long transmission. However, several parameters related to transmission cost need to be considered when selecting a cluster head. While most of the existing research work considers energy and distance as the most stringent parameter to reduce energy consumption, these approaches fail to create a balanced and fair cluster. Consequently, unbalanced cluster formation results in the degradation of overall performance. In this research work, a cluster head selection algorithm is proposed by considering residual energy, number of neighbor nodes, and one hop cluster head information which covers every region of the sensing area in a balanced manner as well as saves significant amount of energy. Furthermore, a capture effect based intra-cluster communication mechanism is proposed that efficiently utilizes the time slot under various traffic conditions. A Naive Bayes Classifier is used to adapt the window size dynamically according to the traffic pattern. Finally, a simulation model using OMNeT++ is developed to compare the performance of the proposed approach with the pioneer clustering approach, LEACH, centralized LEACH-MAC and contemporary DWCA protocol. Simulation results show that proposed approach improved overall performance in terms of energy efficiency, throughput, and network lifetime.