Angle-Based Positioning Estimation Leveraging Diffuse Scattering Paths in Millimeter-Wave MIMO Systems

Position awareness is crucial for various applications in wireless ecosystems. However, adverse propagation and non-line-of-sight (NLOS) conditions, such as those found in complex indoor environments, pose significant challenges for accurate positioning. This paper addresses these challenges by fully utilizing NLOS path information and presents a position estimation method tailored for environments dominated by diffuse reflections without specular reflection components. This method models the problem of determining the number of diffuse signal paths and subspaces via tensor rank estimation and tensor decomposition, respectively. This method exploits the rotational invariance of the subspace to estimate the angle of arrival (AOA) and angle of departure (AOD). Furthermore, this paper models the spatial relationships between base stations (BSs), mobile stations (MSs), and obstacles as spatially directed line segments, moving away from traditional triangulation methods. Additionally, the Cramer-Rao lower bound (CRLB) and positioning error lower bound (PEB) in tensor form are derived for the evaluating the estimation accuracy. The simulation results validate the effectiveness and superiority of the proposed method, and the performance trends are determined under varying obstacle quantities, signal-to-noise ratios (SNRs), and scattering parameters.