Efficient Intelligent Localization Scheme for Distributed Wireless Sensor Networks

Abstract

To estimate location accurately in distributed wireless sensor networks (WSNs), an efficient intelligent position estimation scheme is proposed which needs fewer known nodes without calculating relative positions. The proposed intelligent localization scheme includes two phases. In the first phase, an initialization location algorithm is adopted, where known nodes broadcast their positions and one-hop distances, and then unknown nodes use the received information and operate weighted estimation algorithm to calculate their initial positions. In the second phase, an optimized location algorithm is employed. Unknown nodes exchange information with their neighbors and operate a modified weighted estimation algorithm to refine their positions repeatedly. Additionally, a location scheme also is proposed to solve the problem that a node only has two neighbors. Detailed simulation results show that, the proposed scheme can decrease the average position error down to 9% radio range, and increase the average number of located nodes up to 78% when connectivity is greater than 12 in 20% known node ratio.