Simulated Annealing based Wireless Sensor Network Localization with Flip Ambiguity Mitigation

Abstract

Accurate self-localization capability is highly desirable in wireless sensor networks. A major problem in wireless sensor network localization is the flip ambiguity, which introduces large errors in the location estimates. In this paper, we propose a two phase simulated annealing based localization (SAL) algorithm to address the issue. Simulated annealing (SA) is a technique for combinatorial optimization problems and it is robust against being trapped into local minima. In the first phase of our algorithm, simulated annealing is used to obtain an accurate estimate of location. Then a second phase of optimization is performed only on those nodes that are likely to have flip ambiguity problem. Based on the neighborhood information of nodes, those nodes likely to have affected by flip ambiguity are identified and moved to the correct position. The proposed scheme is tested using simulation on a sensor network of 200 nodes whose distance measurements are corrupted by Gaussian noise. Simulation results show that the proposed scheme gives accurate and consistent location estimates of the nodes and mitigate errors due to flip ambiguities.