AI Heterogeneous Radar Digital Back-end Platform Based on RFSoC

Aiming at the development trend of direct radio frequency sampling in radar systems, we propose an artificial intelligence (AI) heterogeneous radar digital back-end platform based on the RF System-on-Chip (RFSoC). Utilize the RFSOC's radio frequency sampling capacity, the RF signal is directly sampled and synthesized in the proposed system. Specially, the overall system is equipped with accelerated computing scheme based on AI for signal processing, which integrates the Jetson AGX Xavier series GPU to the digital back-end platform. Firstly, the overall box diagram of the system design is proposed. Secondly, the key technologies in the design are discussed. Finally, this platform also provides an important reference value for the construction of next-generation RF sampling radar and radar heterogeneous computing platform.