Direct positioning of a stationary target in MIMO-OFDM system

Abstract

Multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) system, a new concept in recent years, is used extensively in civilian and military applications. We develop a novel direct position determination (DPD) approach in MIMO-OFDM system to locate a stationary target, reaching a superior performance than other conventional localization algorithms. Firstly, we obtain extend noise subspaces from all receivers by constructing and decomposing the extended covariance matrices. Then, the target position is estimated directly via fusing the extend noise subspace data, thereby avoiding the limitations of two-step method. Our proposed algorithm realizes low complexity and high robustness to noise by using the subspace of the extended Hadamard product. Simulation results demonstrate our proposed algorithm outperforms other localization methods, especially under the condition of low signal-to-noise ratio (SNR).