Skip to Main Content
Localization in mobile sensor networks is more challenging than in static sensor networks because mobility increases the uncertainty of nodes' positions. Most existing localization algorithms in mobile sensor networks use Sequential Monte Carlo (SMC) methods due to their simplicity in implementation. However, SMC methods are very time-consuming because they need to keep sampling and filtering until enough samples are obtained for representing the posterior distribution of a moving node's position. In this paper, we propose a localization algorithm that can reduce the computation cost of obtaining the samples and improve the location accuracy. A simple bounding-box method is used to reduce the scope of searching the candidate samples. Inaccurate position estimations of the common neighbor nodes is used to reduce the scope of finding the valid samples and thus improve the accuracy of the obtained location information. Our simulation results show that, comparing with existing algorithms, our algorithm can reduce the total computation cost and increase the location accuracy. In addition, our algorithm shows several other benefits: (1) it enables each determined node to know its maximum location error, (2) it achieves higher location accuracy under higher density of common nodes, and (3) even when there are only a few anchor nodes, most nodes can still get position estimations.