Particle filter for target tracking in multi-modality wireless sensor networks

Most of the target tracking algorithms proposed for wireless sensor networks (WSNs) so far have been relying on sensors of single modality. To integrate multiple sensing modalities (e.g., by using the proximity sensors and ranging sensors together) to improve the tracking performance, the tracking algorithm shall be capable to deal with non-linear and non-Gaussian nature of the tracking problem. In this paper, we present a particle filter algorithm for multi-modality target tracking in WSNs. Simulation results show that the proposed algorithm can provide a good balance among sensor costs and tracking accuracy.