IRS-Assisted Joint Sensing and Communication Design for Autonomous Driving

Abstract

Joint sensing and communication (JSAC) has emerged as a promising technology in autonomous driving, as it allows simultaneous road sensing and two-way communication using a single shared platform. Meanwhile, intelligent reflective surface (IRS) enables sensing enhancement and communication with targets in a blind zone. In this paper, we propose an IRS-assisted JSAC design to address two issues of the limited sensing range of automotive radar and the likely occlusion among road targets. We co-design the IRS’ reflection coefficient vector to steer the beam towards the directions of radar targets as well as embed the communication symbols into the reflected signals. Considering the phase-only property of the passive IRS, we establish a constant modulus co-design problem. We seek to optimize the covariance matrix first and then obtain the optimal reflection coefficient vector via matrix decomposition. Subsequently we transform the constant modulus constraint into a rank-1 semidefinite programing (SDP) problem and solve it iteratively. Simulation results demonstrate the effectiveness of the proposed IRS-assisted JSAC design.