Distributed Massive MIMO FMCW Radar Simulator Based on Ray Tracing

This work introduces a system-level model for radar signal generation, developed to support the latest applications of radar technology like automotive radar industry, high-resolution RF security screening and vital sign monitoring. This study models the dynamic environmental geometry sourced from conventional 3D software such as Blender. Subsequently, it introduces a ray tracing approach incorporating multipath reflections and line of sight based ray allocation to accurately compute the delay and amplitude of received signals based on the radar equation. The proposed model supports a range of conventional multichannel FMCW radar techniques, including phased array, slow time TDM, BPM, DDM MIMO radar for the generation of received signals. The simulation results from the CUDA GPU-accelerated implementation demonstrate the capabilities of the proposed radar signal simulator.