Observation noise modeling based particle filter: An efficient algorithm for target tracking in glint noise environment

In this paper, a novel particle filtering algorithm for target tracking in the presence of glint noise based on observation noise modeling is proposed. The algorithm samples particles using the observation likelihood function, the construction of which is converted to a modeling problem of observation noise. Additionally, the Gaussian mixture model is incorporated to approximate the distribution of observation noise at each time instant. In order to derive a recursive form update for the parameters of the Gaussian components, the maximum likelihood estimation method is employed, enabling noise to be effectively tracked by fusing the latest observations. The algorithm is then used in simulations of bearings-only tracking problems in a glint noise environment with two types of targets: non-maneuvering and maneuvering. The results of the proposed algorithm are evaluated and compared to several existing filtering algorithms through a series of Monte Carlo simulations. The simulation results demonstrate that the proposed algorithm is more precise, robust, and even has a faster convergence rate than the comparative filters. Lastly, the performance of the proposed filter in situations with different numbers of particles and Gaussian components is explored using the simulation results.