All-Weather Automotive Perception: Monte Carlo Simulation of FMCW LiDAR-Radar Sensor Fusion in Foggy Conditions

This work presents an end-to-end Monte Carlo-based framework for FMCW LiDAR-radar fusion in fog, targeting all-weather automotive perception. The model couples radiative-transfer simulation of FMCW LiDAR with a 77 GHz radar link-budget formulation, jointly capturing optical scattering, attenuation, and RF propagation within the same fog microphysical environment. Through integrated beat-spectrum and range Doppler synthesis, the framework quantifies visibility-dependent degradation, clutter evolution, and cross-sensor complementarity. An adaptive fusion kernel balances LiDAR spatial precision with radar penetration as visibility decreases. Simulations over visibilities of 0.1–2.0 km show that while LiDAR alone fails beyond 70 m in dense fog, the fused system sustains >100 m detection at Pd = 0.9, P f a = 10−3. By linking Monte Carlo radiative physics with probabilistic sensor fusion, this study delivers the first quantitative analysis of FMCW LiDAR–radar synergy under scattering-limited conditions, advancing the foundation of all-weather autonomous perception.