Iterative Approaches to Interference Mitigation for Automotive Radar

The problem of interference mitigation for automotive radar becomes increasingly urgent due to the increasing popularity of automotive radar. This paper presents a novel algorithm referred to as interference mitigation via iterative approaches (IMIA) for automotive radar. We first present an iterative constant false alarm rate (CFAR) approach to detect the locations of the strong interferences in the fast-time versus slow-time domain. After nulling the detected interferences, we use the iterative adaptive approach (IAA) to form range-Doppler image. Another CFAR detector is introduced to detect strong targets and refine their parameters using the last step of a relaxation-based cyclic approach (RELAX). After subtracting out the strong targets in the fast-time versus slow-time domain, we can go back to the CFAR detection of the interference locations to detect and nullify weak interferences. This is followed by IAA imaging and target detection and parameter refinement. The aforementioned steps are repeated until no more targets or interferences can be detected. Numerical examples are provided to demonstrate the effectiveness of the proposed IMIA algorithm.