Two-Dimensional Subspace-Based Model Order Selection Methods for FMCW Automotive Radar Systems
In this paper, two novel subspace-based model order selection (MOS) methods to estimate the number of neighboring targets in a 2-dimensional (2-D) radar spectrum are presented. The proposed 2-D MOS methods build on the subspace of the input Hankel block matrix, and could be integrated into most subspace-based high resolution algorithms. Accordingly, in order to adapt the 2-D MOS methods to the automotive radar systems, several signal preprocessing steps are developed. Finally, the methods are evaluated in the framework of an automotive radar system using both simulated data and real data measured by a 77 GHz FMCW radar. The results show that the subspace-based 2-D MOS methods have superior performance than the Akaike Information Criterion (AIC).