PHD Filtering for Multi-Source DOA Tracking With Extended Co-Prime Array: An Improved MUSIC Pseudo-Likelihood
Jun Zhao, Renzhou Gui, Xudong Dong
Abstract
To solve the problem of multi-source direction of arrival (DOA) tracking in co-prime array, a multi-source DOA tracking algorithm based on probability hypothesis density (PHD) filtering is proposed, which can adapt to the scenario where DOA and number of sources change with time. In this letter, we use the minimum description length (MDL) method to estimate the number of sources and construct a new noise subspace by performing eigenvalue decomposition (EVD) on the reconstructed signal subspace. An improved multiple signal classification (MUSIC) pseudo-spectrum is utilized to calculate the likelihood function of the proposed method. The likelihood function is further exponentially weighted to increase the weight of particles. Simulation results show that compared with the existing methods, this algorithm has better tracking performance.