Stochastic Geometry for Radar Detection of VRU and Clutters

This paper introduces a novel approach to Vehicle-to-Vulnerable Road User (V2VRU) communication using stochastic geometry (SG) tools, such as the Poisson Point Process (PPP) and the Neyman-Scott Process, to evaluate the performance of a joint radar system. In this framework, clutters are modeled using the Neyman-Scott Process, while the PPP represents pedestrians. Stochastic geometry provides a robust analytical tool for evaluating radar parameters. Rather than treating each radar as an isolated instance, the SG framework models multiple radars as realizations of a single spatial process. This allows for a unified analysis of parameters such as radar cross section (RCS), signal clutter-to-noise ratio (SCNR), clutter distribution, and detection probability. The results indicate that radar performance under clutter distribution is significantly improved when modeled using the Neyman-Scott Process compared to the PPP.