Latency-Aware NOMA-Powered Joint Communication and Computing Resource Allocation for Industrial IoT Systems

With the rapid development of industrial Internet of Things (IoT) applications, the exponential growth of IoT devices imposes significant challenges on the communication-computing resource allocation, making it difficult to ensure real-time information transmission and processing for latency-sensitive industrial IoT applications. To address this issue, an edge-cloud collaborative computing architecture is established to reduce computational latency by partially offloading tasks from edge servers to the cloud. Under this architecture, non-orthogonal multiple access (NOMA) is adopted to simultaneously access multiple devices on the same spectrum resource. Specifically, a game-theoretic algorithm is proposed to optimize IoT device clustering and the matching between clusters and edge servers. Then, the Hungarian algorithm is utilized for the channel allocation among edge servers. Finally, the bisection search algorithm is adopted to optimize the transmit power to further reduce the transmission delay from IoT devices to edge servers. The proposed scheme achieves efficient communication and computation resource allocation, leading to minimized system-wide computational and transmission latency. Simulation results demonstrate that the proposed scheme outperforms existing schemes in terms of transmission delay and computing delay.