JED-ADMMNet: An ADMM Network-Based Receiver for Uplink ISAC Systems

This letter focuses on uplink integrated sensing and communication (ISAC) systems, where radar-communication base stations assess radar echo signals and simultaneously serve multiple uplink communication users. A receiver design problem is formulated to maximize the joint a posterior probability of radar echo signals and uplink communication signals. To solve this problem, an efficient joint echo signal estimation and communication signal detection (JED) algorithm based on the alternating direction method of multipliers (ADMM) is proposed. To further enhance performance, a model-driven deep learning network called JED-ADMMNet is designed by unfolding the ADMM iterations and incorporating trainable parameters. Numerical results show that the proposed JED-ADMMNet receiver significantly outperform existing receivers in both communication and sensing performance.