On the Partial RSS-Connectivity Based Localization in Wireless Sensor Networks

This paper studies the impact of using partial connectivity information in a received signal strength (RSS) based localization method in wireless sensor networks (WSNs). It aims to find the degree of neighborhood sufficient and necessary to improve localization based on RSS technique. A hybrid approach, based on using jointly the partial connectivity information and the RSS is used. The purposes are evaluating the estimated path loss exponent (PLE) of the propagation model and the positions of unknown nodes. Extensive simulations are conducted to evaluate the normalized mean absolute error (NMAE) of the PLE of the channel using global connectivity and partial connectivity at two different neighborhood levels. Also, cumulative distribution function (CDF) of the NMAE localization compared to Cramer-Rao lower bound (CRLB), using partial and global connectivity are presented. Obtained results show that sufficient accurate estimations can be achieved when exploiting connectivity information at only 2-hops neighborhood, allowing reduce the information exchange.