An Extended Growing Self-Organizing Map for Selection of Clusters in Sensor Networks

Abstract

Sensor networks consist of wireless enabled sensor nodes with limited energy. As sensors could be deployed in a large area, data transmitting and receiving are energy consuming operations. One of the methods to save energy is to reduce the communication distance of each node by grouping them in to clusters. Each cluster will have a cluster-head (CH), which will communicate with all the other nodes of that cluster and transmit the data to the remote base station. In this paper, we propose an extension to Growing Self-Organizing Map (GSOM) and describe the use of evolutionary computing technique to cluster wireless sensor nodes and to identify the cluster-heads. We compare the proposed method with clustering solutions based on Genetic Algorithm (GA), an extended version of Particle Swarm Optimisation (PSO) and four general purpose clustering algorithms. This could help to discover the clusters to reduce the communication energy used to transmit data when exact locations of all sensors are known and computational resources are centrally available. This method is useful in the applications where sensors are deployed in a controlled environment and are not moving. We have derived an energy minimisation model that is used as a criterion for clustering. The proposed method can also be used as a design tool to study and analyze the cluster formation for a given node placement.

Keywords:

• energy optimisation
• Self-Organizing Map
• evolutionary computing
• sensor networks.