REDTS: RLS-Enhanced Doppler Target Separation for Automotive FMCW Radar Systems

This paper proposes an advanced interference suppression and target separation technique for automotive Frequency Modulated Continuous Wave (FMCW) radar systems. By employing Recursive least squares (RLS) adaptive filters, the method effectively separates moving and stationary targets while suppressing interference from various sources, such as FMCW, Range Division Multiple Access (RDMA), and Code Division Multiple Access (CDMA) signals. The approach, named RLS-Enhanced Doppler Target Separation (REDTS), uses Doppler frequency differences to isolate targets with different motion states and processes these signals in continuous time, ensuring high accuracy and real-time performance in complex environments. Simulations and real-world measurements demonstrate the method robustness, exhibiting significant improvements in target detection accuracy even when trained for different scenarios than the ones tested. Our interference suppression proposal enhances the reliability of automotive radar systems, allowing for efficient target separation with minimal computational overhead.