Precoding and Bi-Static Sensing for Cooperative ISAC

Abstract

Inter-cell interference and reflections are typically treated as detrimental effects in integrated sensing and communication (ISAC) systems, despite their potential to provide additional sensing information in multi-cell networks. This paper proposes a precoding and bi-static sensing framework for cooperative ISAC in multi-user multiple-input multiple-output networks. Leveraging the cooperative paradigm of overlay cognitive radio (CR), the primary system shares channel state information and user data with the cognitive system, enabling simultaneous communication and sensing over the same spectrum. The cognitive system performs integrated sensing and downlink transmission using symbol-level precoding that exploits both intra- and inter-cell constructive interference (CI), while inter-cell reflections are utilized as additional sensing sources to enable bi-static target estimation. The transmit symbols are optimized by minimizing the Cramér-Rao bound (CRB) under CI-based communication and per-antenna power constraints. The resulting optimization problem is formulated using semi-definite relaxation (SDR) and solved via convex optimization. To improve scalability, a lowcomplexity successive convex approximation (SCA) algorithm is also developed, achieving performance close to that of the SDR solution with significantly lower computational complexity. Numerical results show that the proposed bi-static design substantially improves angle estimation accuracy compared with monostatic sensing and mean-square-error (MSE)-based beamforming approaches.