Performance Evaluation of a Hybrid RSS-SOM Localization Algorithm for Wireless Ad-Hoc Networks

摘要

Localization of wireless ad hoc networks has gained much research attentions for several years. This paper proposes a hybrid localization algorithm which exploits Received Signal Strength (RSS)-based ranging and Self Organizing Maps (SOM)-based range free localization methods to obtain the tradeoff between cost, power and location accuracy. Distance information from RSS measurement has been utilized in the learning steps of SOM-based localization algorithm to get more accurate location estimates while reducing number of learning steps. Method on RSS uncertainty reduction is also incorporated in the proposed hybrid RSS-SOM algorithm. Results from extensive simulations prove that the hybrid RSS-SOM algorithm outperforms several existing localization algorithms for all node density, anchor utilization and the number of learning steps.