Real-time Signal Processing System Development for Automotive Cascaded Radar

Abstract-This paper presents the development of a real-time signal processing system for automotive radar, addressing the increasing demand for advanced driver assistance systems (ADAS) to enhance driving experiences and safety. Radar, as a key sensor in ADAS, enables reliable object detection and tracking, even in challenging weather conditions. By employing cascaded integration of radar transceiver MMICs (Monolithic Microwave Integrated Circuits), the system achieves an expanded number of transmitter (TX) and receiver channels (RX), resulting in improved radar performance, including enhanced range resolution, sensitivity, and angular resolution. The paper focuses on the design of the real-time signal processing system, utilizing cascaded MMICs and a dedicated MCU for radar processing. Synchronization mechanisms are implemented to ensure precise timing control of the cascaded MMICs, while the system employs a real-time signal processing architecture with three CPUs and two hardware accelerators in parallel. Performance verification confirms the synchronized operation of the cascaded MMICs within a 1.4ns timing window, enabling real-time updates of up to 512 targets every 50ms with consistent delay.