Source Localization for Extremely Large-Scale Antenna Arrays with Spatial Non-Stationarity

Extremely large-scale antenna array (ELAA) is a promising technique in 6G and autonomous driving thanks to its high spatial resolution. However, due to the extremely large array aperture, the sources may only "see" a portion of the array, called visibility region (VR). Since the information of VR is usually unknown a priori, traditional methods may encounter performance degradation. In this paper, we show that under the exact steering vector model of ELAA, the eigenvectors of the signal subspace and the steering vectors are approximately collinear in most scenarios. Thus, the angle and range can be easily estimated by using the eigenvectors of the signal subspace. Moreover, the VRs of each source can be also estimated from the eigenvectors of signal subspace. Numerical results show that the estimation performance of our method is comparable to MUSIC with known VR information.