Genetic Resampling Particle Filter with Randomly Delayed Measurements and Its Application to Line-of-Sight Rate Estimation

Abstract

To deal with randomly delayed measurements and glint noise, a novel recursive filter referred as a randomly delayed genetic resampling particle filter (RD-GRPF) is proposed in this paper. By making use of Bernoulli random variables, the measurement model is modified to describe the random delay. Then, the weight update equation is reformulated based on this model. To avoid the particle degeneration and sample impoverishment that always arise in the application of a particle filter, the genetic resampling method is utilized to resample the particles. Then, the RD-GRPF is obtained. Therefore, the filter proposed not only copes with randomly delayed measurements, it also keeps the advantage of the standard particle filter, which ensures good performance in the case of non-Gaussian noise. In addition, the RD-GRPF proposed is applied to line-of-sight (LOS) rate estimation, the model for which is also presented in this paper. A simulation was conducted and the results demonstrate the superiority of the RD-GRPF.