A Computationally Efficient Radar Sensor Modeling Methodology Incorporating Rain-Induced Propagation Attenuation for Virtual Validation of Autonomous Driving Systems

Abstract

This paper presents a computationally efficient radar sensor modeling methodology incorporating rain-induced propagation attenuation for simulation-based validation of autonomous driving systems. The proposed framework combines the classical radar equation with the rain attenuation model specified in ITU-R Recommenda-tion P.838-3 to estimate received signal power degradation and target-range estimation errors under rainfall conditions. A stochastic False Alarm Probability Generation (FAPG) algorithm is also introduced to reproduce intermittent tar-get detection according to user-defined probabilities of false alarm and detection. The proposed methodology was implemented through real-time co-simulation between MATLAB/Simulink and IPG CarMaker. Simulation results demonstrated that the proposed framework successfully reproduced rain-induced propagation attenuation, target-range estimation errors, received signal power degradation, and stochastic target detection behavior while maintaining computational efficiency suitable for real-time simulation.