Symbol-Level Precoding for Minimizing Ranging CRB in ISAC Systems: A Low-Complexity Approach

Abstract

This letter investigates a dual-functional waveform design for MIMO-OFDM integrated sensing and communication (ISAC) systems. Firstly, an optimization problem of minimizing the radar ranging Cramér-Rao bound (CRB) is formulated under the constraints of multi-user communication quality of service (QoS) and transmit power. To fully exploit the spatiotemporal degrees of freedom (DoFs), a novel symbol-level precoding (SLP) scheme is designed to recast the multi-user interference into constructive interference (CI). To tackle the resulting NP-hard non-convex fractional optimization problem, a low-complexity algorithm based on quadratic transform and alternating optimization (QT-AO) is proposed. Simulation results demonstrate that the proposed SLP-ISAC scheme achieves a radar ranging CRB up to 4 dB lower than the conventional semidefinite relaxation (SDR)-based block-level precoding (BLP) under heavy communication loads, along with significantly faster convergence rate and reduced computational complexity.