Reconfigurable Beamforming for Automotive Radar Sensing and Communication: A Deep Reinforcement Learning Approach

In this article, we present a novel low-cost, dual-function radar-communication system that addresses dynamic environments such as those arising in automotive applications. The low cost is achieved by using a sparse phased arrays equipped with quantized double-phase shifters. The operation in dynamic environments is achieved via a deep reinforcement learning (DRL) approach that adaptively selects a small subset of transmit antennas and adjusts the phase shifters such that the transmitted energy is concentrated on the communication user and the target of interest, while the interference to other radars is reduced. The action space in the DRL approach increases fast with the number of antennas and the number of bits used in quantization, and as a result the complexity of the design problem grows exponentially. To tackle the resulting curse of dimensionality in the action space, we adopt the Wolpertinger strategy, which incorporates the nearest neighborhood component to project the vast action space into a smaller, more manageable space while maintaining the desired performance. Numerical results demonstrate the feasibility of our proposed method.