TDLoc: Passive Localization for MIMO-OFDM System via Tensor Decomposition

Passive localization is an important aspect of integrated sensing and communication (ISAC). However, it is challenging to estimate the target position and velocity accurately from the receiving signals due to complex multipath propagation. This article presents TDLoc, a multiple-input-multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM)-based passive localization and tracking system, using channel state information (CSI). We first proposed a fourth-order tensor model that contains the Angle-of-Departure (AoD), Angle-of-Arrival (AoA), Time-of-Flight (ToF), and Doppler frequency shifts (DFSs) information, followed by developing an efficient joint estimation algorithm. We also show that with more than one pair of transceivers, our method can obtain the target velocity from the relativistic Doppler effects, leading to additional DFS-based trajectory information. Moreover, the Cramer-Rao lower bound (CRLB) for multipath parameter estimation and positioning is derived for performance evaluation. Numerical results show that TDLoc outperforms state-of-the-art methods in terms of localization accuracy.