Enhancing Massive Access in ISAC Networks: A Non-Orthogonal Random Access Scheme With Semi-Grant-Free Strategy

Integrated sensing and communication (ISAC) is viewed as a feasible framework for the sixth-generation (6G) systems, which has the advantages of efficient spectrum utilization and low hardware cost. Non-orthogonal multiple access (NOMA) is just an effective method for meeting the requirements in ISAC due to its efficient interference management and flexible resource allocation. However, this kind of conjunction would be disrupted when encountering the scenarios of massive machine-type communication devices (MTCDs), especially for devices with strict delay quality of service (QoS) requirements. To fill aforementioned imperfection, we propose a non-orthogonal random access scheme with semi-grant-free strategy (SGF-NORA) to improve throughput for ISAC networks with massive MTCDs. Specifically, the two-step early operation for the proposed NORA scheme includes access control and tagged preambles, which control the average access number of devices and avoid inter-NOMA power collisions. Firstly, we abstract a model of uplink ISAC networks with non-orthogonal random access, and further derive the achievable rate of short packet transmissions and radar estimation information rate (REIR). Relying on the martingale theory, we present an elegant delay analysis for each MTCD in ISAC networks with SGF-NORA scheme and establish a throughput maximization problem under the dual constraints of sensing and communication. To tackle the complex coupled optimization problem, we decouple it into sub-problems of access control and power allocation, which are solved by the zebra optimization algorithm (ZOA) and successive convex approximation (SCA), respectively. Furthermore, a joint access control and power allocation (JACPA) algorithm is devised to implement resource allocation. Simulation results demonstrate that the SGF-NORA solution outperforms benchmark schemes and the JACPA algorithm offers competitive performance.