Target localization and Doppler Estimation Using Automotive MIMO Radar

In this paper, a joint parameter estimation method based on tensor decomposition is introduced, aimed at estimating multiple parameters of moving targets with automotive time-division multiplexing multiple-input multiple-output (MIMO) radar utilizing frequency modulated continuous wave (FMCW) signals. To overcome the constraints of automotive radar systems imposed by the own spatial size of vehicle, a virtual array with high degrees of freedom is constructed through the sum co-array formed by the synthetic aperture technology based on vehicle motion. By reassembly of received observations, we construct a four-dimensional tensor, organizing the echo data within a framework that integrates range, velocity, and angular information. Furthermore, in order to prevent information loss associated with the process of tensor decomposition, parallel factor analysis (PARAFAC) is applied. We integrate PARAFAC with alternating least squares optimization to perform dimensionality reduction on multidimensional data, providing convenience for the subsequent estimation of parameters for each dimension. Simulation results validate the dependability of the proposed algorithm.