RCS Measurement and Modeling of Typical Traffic Participants Based on Millimeter-Wave Radar for Autonomous Vehicles

At present, 77GHz millimeter-wave (MMW) radar has become a critical sensor in intelligent transportation systems due to its all-weather detection capability, which enables it to resist complex weather and light interference. Radar cross section (RCS) is a significant characteristic of radar, greatly impacting the detection quality of traffic targets across various traffic scenarios. RCS is usually measured in an anechoic chamber to establish a model of the RCS of typical traffic participants. However, due to large target fluctuations and multi-angle scattering centers of targets, representing the RCS characteristics of typical traffic participants with a single point is challenging. Taking global vehicle target (GVT), pedestrian target and cyclist target as examples, this paper proposes a method for measuring and modeling the RCS features of typical traffic participants. For the static RCS features of targets, we measured the RCS of the target under different viewing angles in an anechoic chamber, establishing an RCS feature analysis method for the target based on the radar viewing angle. For the dynamic RCS features of targets, we collected radar reflection data of typical traffic participants from different perspectives while in motion in an open-field environment. We then extracted and synthesized the strong scattering centers of the target, established a dynamic RCS feature model, and fitted the RCS boundary curve from the near field to the far field. Based on the road scenario, we established the RCS simulation scenario to validate the effectiveness of the target dynamic RCS simulation model in the road scenario. The results of testing and analysis show that radar enables stable characterization and modeling of target RCS, thereby enhancing the environmental perception capabilities of autonomous vehicles and improving driving safety within intelligent transportation systems.