Node-Replacement Policies to Maintain Threshold-Coverage in Wireless Sensor Networks

• Published in: 2007 16th International Conference on Computer Communications and Networks pp. 760 - 765
• Main Authors: Parikh, S., Vokkarane, V.M., Liudong Xing, Kasilingam, D.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.08.2007
• Subjects: Application software; Biosensors; Internet; IP networks; Quality of service; Radar tracking; sensing coverage; Sensor phenomena and characterization; Sonar detection; Target tracking; Wireless sensor networks
• ISBN: 9781424412501; 1424412501
• ISSN: 1095-2055
• DOI: 10.1109/ICCCN.2007.4317909

With the rapid deployment of wireless sensor networks, there are several new sensing applications with specific requirements. Specifically, target tracking applications are fundamentally concerned with the area of coverage across a sensing site in order to accurately track the target. We consider the problem of maintaining a minimum threshold-coverage in a wireless sensor network, while maximizing network lifetime and minimizing additional resources. We assume that the network has failed when the sensing coverage falls below the minimum threshold-coverage. We develop three node-replacement policies to maintain threshold-coverage in wireless sensor networks. These policies assess the candidature of each failed sensor node for replacement. Based on different performance criteria, every time a sensor node fails in the network, our replacement policies either replace with a new sensor or ignore the failure event. The node-replacement policies replace a failed node according to a node weight. The node weight is assigned based on one of the following parameters: cumulative reduction of sensing coverage, amount of energy increase per node, and local reduction of sensing coverage. We also implement a first-fail-first-replace policy and a no-replacement policy to compare the performance results. We evaluate the different node-replacement polices through extensive simulations. Our results show that given a fixed number of replacement sensor nodes, the node-replacement policies significantly increase the network lifetime and the quality of coverage, while keeping the sensing-coverage about a pre-set threshold.