DOA Estimation for a Mixture of Completely and Partially Polarized Signals With Enhanced Separation
Yujian Pan, Xin Xu, Afeng Yang · IEEE Transactions on Aerospace and Electronic Systems · 2024
The direction-of-arrival (DOA) estimation for a mixture of completely polarized (CP) and partially polarized (PP) signals is essential, yet it has not been thoroughly explored. This paper presents a practical method with enhanced separation for this problem. First, the source enumeration of CP and PP signals is resolved via conducting model order estimations on the array covariance matrix and a smoothed covariance matrix with no threshold needed. Then, by the utilization of feature differences between CP and PP signals, a PP-signal-sensitive cost function free of the conventional issue of CP signal leakage is proposed to exclusively estimate the DOAs of PP signals. Lastly, the oblique projection operator (OPO) is introduced to extract the data without PP components, upon which the DOA estimation for CP signals is performed and the previous problem of CP signal invisibility is avoided. To estimate the OPO, an iterative method and a non-iterative method are proposed, both of which have higher accuracy than the existing methods and remain effective under high dynamic range of inputs signals. Moreover, the grid-search has been circumvented by leveraging the root-MUSIC algorithm. Simulation results demonstrate the superior performance of the proposed method compared to the state-of-the-art methods.