Gaussian particle filtering algorithm for out-of-sequence-measurement problem

Using multiple sensors to track targets, due to communication delays and variable signal pre-processing time, will lead to out-of-sequence measurement problems. In this paper, based on sequential Bayesian formula and Gaussian particle filter, a novel out of sequence measurement processing algorithm is developed. In essence, this algorithm uses importance sampling to update the posterior means and their covariances, and also approximates the posterior distributes by single Gaussians. The algorithm has low complexity and its performance is consistent with standard sequential processing algorithm and it is asymptotically optimal as numbers of particles tends to infinity. This paper is the extended version of [15].