Energy Efficient Cluster based Mobility Prediction for Wireless Sensor Networks

Wireless Sensor Networks (WSN) is a collection of wireless sensor nodes forming a temporary network without using any established infrastructure or centralized administration. Sensor nodes are battery operated and may be mobile, critical issues is to reduce the energy conservation and prediction of node mobility to maximize the network lifetime by using limited network resources. In order to overcome these issues, in this paper we propose an Energy Efficient Cluster Based Mobility Prediction Routing Protocol (EECMPR) for WSNs using Kalman Filters. In this technique, the node with high weight is chosen as cluster head based on the parameters such as residual energy and transmission range. The cluster head collects the information from the nodes in its cluster and transmits the information to the sink. The position of each mobile sensor is carried at the respective nodes using Kalman filter which gives the State Update which consists of velocity, acceleration and position. This State Update is sent to the cluster head which will process the information and in turn will send it to sink. Before receiving the information from the cluster head, the sink would do prediction and compare the received results obtained from cluster head and the predicted result. Finally it will send data packets to the nodes after determining the position. By simulation results, we have shown that the proposed technique (EECMPR) minimizes the overhead and improves accuracy in predicting the position of the node.