Acoustic source localization in noisy wireless sensor networks

Target localization is one of the typical applications of wireless sensor networks. However, there are few localization algorithms addressing the problems caused by both the noisy wireless channels and the limited sensing range and energy of sensor nodes. In this paper, a new source localization algorithm based on maximum likelihood (ML) estimation incorporating wireless channel impact is proposed. The Cramer-Rao Bound (CRB) of the ML acoustic source location estimate has been derived to analyze the accuracy of location estimations. Extensive simulations have been conducted. The results show that this proposed ML method can reach its CRB performance bound with relatively small amount of data and consistently outperforms the source localization methods without considering the channel impact.