Optimizing Task Offloading for Power Line Inspection in Smart Grid Networks with Edge Computing: A Game Theory Approach

Abstract

In the power grid, inspection robots enhance operational efficiency and safety by inspecting power lines for information sharing and interaction. Edge computing improves computational efficiency by positioning resources close to the data source, supporting real-time fault detection and line monitoring. However, large data volumes and high latency pose challenges. Existing offloading strategies often neglect task divisibility and priority, resulting in low efficiency and poor system performance. This paper constructs a power grid inspection offloading scenario using Python 3.11.2 to study and improve various offloading strategies. Implementing a game-theory-based distributed computation offloading strategy, simulation analysis reveals issues with high latency and low resource utilization. To address these, an improved game-theory-based strategy is proposed, optimizing task allocation and priority settings. By integrating local and edge computing resources, resource utilization is enhanced, and latency is significantly reduced. Simulations show that the improved strategy lowers communication latency, enhances system performance, and increases resource utilization in the power grid inspection context, offering valuable insights for related research.