Using Multi-agent System for Solving Coverage Problem in Wireless Sensor Network

Abstract

Wireless sensor network (WSN) is one of the most important paradigms in computer networks because of the widespread applications. Coverage problem is a fundamental issue in sensor networks that reflects how the network is controlled by the sensors, this problem appears when any node becomes failure or out of the range, in this case the area will be disconnected and the data will not send to the destination. We present a new approach which uses a multi-agent system to solve this problem and perform an easy and secure network. In order to do that we implement sensor network by four phases: first construct a virtual network by matlab, second we use k-means clustering to cluster nodes in k-groups, third put the intelligent sensor in each cluster to be as a head for its group, fourth we divide the network to four regions and the closet agent to the sink will be the delegate to send the aggregated data from its region to the destination. Therefore, we tried to minimize the power consumption in WSN, we save the energy by keeping it sleep until it has a task to do , at this case the node changes its status to be in active mode and when it finishes it will be idle.