The Complete Pareto Points for In-band Full Duplex Integrated Sensing and Communication Systems

In this paper, we investigate beamforming and transmit power optimization for in-band full-duplex integrated sensing and communications (FD-ISAC) systems. Our focus is on an FD-ISAC base station (BS) that simultaneously communicates with both downlink (DL) and uplink (UL) communication users (CUs) while detecting a target. Existing FD-ISAC studies typically: (i) consider either DL or UL users, not both, leading to additional interference, (ii) neglect radar signal-to-interference-plus-noise ratio (SINR) as a performance metric, or (iii) address single-objective optimization, focusing on either communication or radar performance. We formulate the problem as a weighted multi-objective optimization, balancing the achievable sum rate for CUs and the received radar SINR-two inherently conflicting objectives. To solve this non-convex problem, we propose two solutions: A complexity-oriented design (COD) utilizing convex transformations and relaxations. And a performance-oriented design (POD) leveraging single-objective solutions to address the multi-objective formulation. Numerical evaluations demonstrate that both methods achieve comparable performance for CUs and radar. However, COD performs better for radar, while POD is superior for CUs, especially with a higher number of BS antennas. COD also has lower computational complexity. Our proposed FD-ISAC schemes outperform existing half-duplex (HD) ISAC schemes by approximately 6 dB for radar SINR and 8 bits/s/Hz for CU rate.