Particle filter based sensor selection in binary sensor networks

This paper is concerned with the application of target tracking in a network of sensors that provide binary output. The binary sensor network tracking problem is formulated in a sequential Bayesian estimation framework and is readily solved by means of a particle filter. We will perform sensor selection by means of a newly proposed scheme. This proposed approach is especially suitable for problems that exhibit non-Gaussian and multi modal behavior in the a posteriori density. We will show that the newly proposed selection scheme can also be extended to the multi target case. In addition, we also formulate a fairly simple sequential approximation of the proposed scheme that dramatically reduces the size of the associated optimization problem.