Joint power distribution and direction of arrival estimation for wideband signals using sparse Bayesian learning
Yi Wang, Baixiao Chen, Yisong Zheng, Minglei Yang · IET Radar Sonar & Navigation · 2016
In this study, a novel algorithm called wideband sparse Bayesian learning (WSBL) is proposed for the power distribution and the direction of arrival (DOA) estimation of wideband signals. The received signals are firstly converted into the time–frequency domain by discrete Fourier transform, then the wideband DOA estimation problem can be constructed as a special multiple measurement vectors case of the sparse signal recovery problem. The WSBL algorithm, which exploits the source power distribution at different frequency bins is proposed for the sparse spectral estimation. It is shown that WSBL has good performance in low signal‐to‐noise ratio and small snapshot number scenario. Without the prior knowledge of source number and focusing matrices, WSBL can estimate the source power distribution and the DOAs of wideband signals simultaneously. In addition, WSBL can also avoid ambiguity even when the array does not meet the half‐wavelength spacing condition. The sensor spacing can be larger than the half‐wavelength at the lowest frequency of the wideband signals. Simulation results show the efficiency of the proposed method.