Sequential Monte Carlo localization in mobile sensor networks

Node localization in wireless sensor networks is essential to many applications such as routing protocol, target tracking and environment surveillance. Many localization schemes have been proposed in the past few years and they can be classified into two categories: range-based and range-free. Since range-based techniques need special hardware, which increases the localization cost, many researchers now focus on the range-free techniques. However, most of the range-free localization schemes assume that the sensor nodes are static, the network topology is known in advance, and the radio propagation is perfect circle. Moreover, many schemes need densely distributed anchor nodes whose positions are known in advance in order to estimate the positions of the unknown nodes. These assumptions are not practical in real network. In this paper, we consider the sensor networks with sparse anchor nodes and irregular radio propagation. Based on Sequential Monte Carlo method, we propose an alterative localization method-Sequential Monte Carlo Localization scheme (SMCL). Unlike many previously proposed methods, our work takes the probabilistic approach, which is suitable for the mobile sensor networks because both anchors and unknown nodes can move, and the network topology need not be formed beforehand. Moreover, our algorithm is scalable and can be used in large-scale sensor networks. Simulation results show that SMCL has better localization accuracy and it can localize more sensor nodes when the anchor density is low. The communication overhead of SMCL is also lower than other localization algorithms.