Agent Based Localization Using Mobile Sink in Wireless Sensor Networks

Abstract

This paper presents agent based localization using the mobile sink in wireless sensor networks. This proposed scheme is less expensive than GPS. It has a longer life time and more energy-efficient. The mobile sink has a large and easily replenishable energy reserve and is movable within the sensor network's deployment area. In this work, a mobile sink traverse inside the network's boundary and gathers information at low energy cost. The use of a mobile sink in localization introduces a new way to reduce energy consumption in WSNs. The location of a mobile sink beacon signal is known to all the sensor nodes and is also communicated periodically to all the sensor nodes. The distance measurements of the three beacon signals broadcasted by the mobile sink moving in a predetermined path and time slot are considered in this scheme, which uses the tri-lateration method to compute the position of the node. For isolated nodes, location estimation is performed for non-GPS equipped nodes to derive from the network their locations by using the reference node beacon signals and performing multilateration. For nodes that receive only one beacon signal from the sink, position estimation is performed by considering the reference node beacon signal by iterative multilateration technique. In this scheme reference nodes are those nodes that are localized by the sink and it does not include any GPS. To test the proposed algorithm the proposed work is simulated using C and some of the performance parameters used for the evaluation of the scheme are localization time, localization error, data gathering time, and communication overhead. The proposed work is compared with the centralized algorithm in which the proposed work outperforms compared to the existing work.