An optimal importance sampling based particle filtering for channel parameter estimation in shallow ocean

Abstract

Estimating channel parameters in a shallow ocean environment is challenging due to low signal-to-noise ratio (SNR), multi-path effect and time-varying nature of ocean. In this paper, a Bayesian framework and its particle filtering (PF) implementation are introduced to cope with this problem. At each time step, the particles are sampled according to a random walk model, and then evaluated by the corresponding importance weights. An extended Kalman filter (EKF) is incorporated to achieve an optimal importance sampling, by which the states are coarsely estimated and the particles are relocated. As such the particles are more likely drawn at the relevant area and can be resampled more efficiently. Experiments show that the proposed EKF-PF tracking algorithm significantly outperforms the traditional tracking approaches in challenging environments.