Optimal Sparse MIMO Transceiver Design for Joint Automotive Sensing and Communications

Joint automotive sensing and communication assisted by optimal sparse MIMO transceiver design is a promising technology for autonomous driving as it reduces hardware cost while preserving high angular resolution. In this paper, we propose to co-design a shared sparse MIMO transceiver within the paradigm of joint sensing and communication (JSAC). Antenna selection is performed to minimize the Cramer–Rao bound (CRB) for accurate tracking with enhanced direction of arrival (DOA) estimation. Meanwhile, the spatial precoding matrix for communications, which exhibits the same sparsity structure with the shared transmitter for automotive sensing, is optimized to deliver a desired quality of service. A solution of this problem requires the application of a series of convex relaxation strategies to transform the resultant non-convex co-design problem into a convex form. The fractional inequality is transformed into the denominator inequality with a constrained numerator and reweighted l1-norm minimization is utilized to promote binary sparsity. Simulations are provided to demonstrate the effectiveness of the optimal sparse MIMO transceiver obtained by the proposed method.