Joint Source Number Detection and DOA Estimation via Reversible Jump MCMC

In this paper, a new Bayesian array signal model structure based on signal reconstruction is proposed, that allows us to define a posterior distribution on the parameter space, which is applicable to both wideband and narrowband signals. Unfortunately, a direct evaluation of this distribution and of its features, including posterior model probabilities, requires evaluation of some complicated high-dimensional integrals. We develop an efficient hybrid sampling algorithm based on reversible jump Markov Chain Monte Carlo methods to jointly detect and estimate the sources impinging on a single array of sensors in Gaussian noise. This algorithm provides higher resolution than traditional wideband methods. The accuracy of the structure and the validity of this method are well verified by the simulation.