Sparse Bayesian Learning for Off-Grid DOA Estimation in Alpha Noise
Wenchao He, Longkai Liang, Lv Xuhao · 2021 China Automation Congress (CAC) · 2021
Sparse Bayesian (SBL) method is widely used in DOA estimation recent years. This method can obtain high performance with fewer snapshots. However, it depends on the probability density of noise, which usually assumed to be Gauss distribution. And the performance of this method reduces in alpha noise which do not have a probability density of closed-form expression. In this paper, we proposed an algorithm which can deal with this situation using SBL algorithm. To reduce the computational complexity of SBL algorithm, an off-grid method is proposed. Simulation results show that algorithm can obtain high performance in alpha noise efficiently.