A Lyapunov-Based Resource Allocation Method for Edge-Assisted Industrial Internet of Things

In edge computing-enhanced industrial Internet of Things (IIoT) environments, networks with multiple terminal equipments (TEs) and edge servers (ESs) face significant challenges in energy-efficient task offloading and resource allocation due to dynamic computational demands and uncertain energy supply. We propose a novel system framework that organizes computation tasks and energy into queues for each TE, allowing TEs to acquire energy and offload tasks to adjacent ESs. This framework optimizes key variables, such as transmission power, CPU processing speeds, and data volume to reduce latency and improve energy efficiency. Additionally, we introduce a Lyapunov optimization-based coalition game approach for computation offloading and resource allocation. This method effectively addresses the coupled and nonconvex optimization challenge by ensuring the system stability and efficiently balancing delay and energy consumption. Comprehensive simulation experiments have shown that this algorithm significantly reduces the cost of ES-Assisted IIoT systems and improves the system performance.