An over-the-horizon radar multipath fusion algorithm

Sky wave over-the-horizon radar (OTHR) usually suffers from multipath propagation effects and low detection probabilities when tracking multiple targets. In this paper we present a multipath unscented Kalman-Gaussian mixture probability hypothesis density (MPUK-GMPHD) fusion framework to overcome these defects. The GMPHD filter is capable of avoiding data associations while maintaining low computational complexity. A probability hypothesis density fusion algorithm is proposed for OTHR, wherein multipath propagation effects are analyzed using a multi-sensor model. Particularly, the target number over-estimation problem is effectively avoided by fusing the multipath information, and an unscented Kalman filter is utilized to solve the nonlinear problem in measurement model. Simulation results show that the proposed MPUK-GMPHD framework estimates target state and number more accurately than conventional methods in complex environments, e.g. OTHR multi-target tracking. It overcomes the target number over-estimation and high computational complexity problems in existing GMPHD filter based algorithms.