Efficient DOA and Magnitude Estimation Jointly Utilizing GAMP and Reweighted LASSO
Shiqi Shu, Ye Tian, Hongyun Zhao, Junying Ren · 2024
Most of existing direction-of-arrival (DOA) estimation methods are investigated under the scenarios of multiple measurement vectors (MMVs), which could be invalid or suffer from severe performance degradation when only single measurement vector (SMV) is available. Meanwhile, it is worth emphasizing that they have not yet focused on signal magnitude estimation, but magnitude information actually has important applications in wireless communications and various industrial fields. Under such a circumstance, an efficient DOA and magnitude estimation method is proposed in this paper, where a two-step estimation scheme is adopted. The first step is to obtain DOA estimation via the generalized approximate message passing (GAMP) algorithm; with available DOAs, the signal magnitudes are then achieved at the second step by exploiting the reweighted least-absolute selection and shrinkage selection operator (LASSO). Simulation results show that the proposed can yield a satisfactory result in a computationally efficient manner.