Article ID: 2024EAP1130
The integration of numerous IoT devices into the distribution network supports coordinated control of grid-connected devices, but the complex topology of distribution networks and uneven base station distribution result in weak coverage and uneven load, leading to poor end-to-end latency performance. While IoT heterogeneous integrated networking technologies enable low-latency access for many power devices, challenges like communication resource competition and slow optimization under uncertain network conditions remain. To address these issues, this paper proposes a joint optimization model for relay scheduling, data compression, and time scheduling, aiming to minimize average end-to-end latency. A two-stage edge-end cooperative resource optimization algorithm based on Lyapunov optimization theory is proposed. In the first stage, a relay scheduling algorithm using relay device connectivity and queue delay-aware ascending price matching optimizes scheduling by dynamically adjusting channel matching costs based on connectivity and queue backlogs. The second stage introduces a delay deviation-aware adaptive particle swarm optimization to optimize time scheduling and data compression, achieving fast convergence. The relay scheduling preferences are updated based on the final objective function value. Simulation results demonstrate the method's effectiveness in reducing latency, improving network performance, and efficiently utilizing network resources.
