Waveform Design for Joint Radar-Communication Systems in Urban Micro-cells

As the demand for spectral efficiency continues to increase, Joint Radar Communication systems have been identified as a critical component of future 6th Generation wireless networks, particularly within Urban Micro-cells. This environment poses significant challenges, including multi-path fading, considerable shadowing, and the need for high-resolution sensing of dynamic objects such as pedestrians and vehicles. This research aims to optimize an integrated waveform, addressing the trade-off between high data-rate communication and high-resolution radar sensing capabilities. By optimizing the Ambiguity Function for sensing and Spectral Efficiency for communication, this optimized waveform aims to minimize cross-interference between systems, thus achieving maximum utilization of resources. By leveraging a Deep Multi-Agent Reinforcement Learning(MARL) framework and the algorithm of Twin Delayed Deterministic Policy Gradient (TD3), this research aims to dynamically optimize the phase shifts of Reconfigurable Intelligent Surfaces (RIS) to improve waveform beamforming. This study has shown significant findings, proving the optimized waveform’s ability to achieve accurate sensing capabilities within a highly cluttered urban environment, meeting low-latency requirements for micro-cellular communication systems. The simulation data has confirmed a Detection Probability (pd)of 0.95 at a Signal-to-Noise Ratio of 10 dB, as well as a 15% reduction in Root Mean Square Error (RMSE) for radar estimation compared to standard OFDM systems.