Energy-Efficient Node Selection for Acoustic Source Localization in Wireless Sensor Network

Abstract

An Energy-efficient Node Selection (ENS) mechanism has been proposed to improve the localization precision of target tracking with low computational complexity and energy consumption. The localization is based on wireless multimedia sensor network with cellular topology. The ENS mechanism utilizes the Global Node Selection algorithm in Energy-based acoustic source localization (GNSE) and the Closest Node Selection (CNS) method synthetically. Firstly the cluster head simply uses the CNS method to select nodes. And then it decides whether to use the GNSE algorithm via evaluating the determinant condition with the information of nodes selected by the CNS. By transmitting a packet only to the selected nodes, instead of all the target-sensed ones, to ask for the position and measurement information, energy could be saved. The simulation shows that the ENS is 67.61% more precise than the CNS, and saves 74.07% energy relative to the GNSE and the CNS when the computational energy consumption is ignored. The ENS mechanism achieves obtaining high localization precision with fewer nodes and prolongs the lifetime of the network greatly.