Cache-Aware Association Strategies for Energy-Efficient MEC-Enabled IIoT Networks

Multi-access edge computing (MEC) has emerged as a key enabler for meeting the stringent performance demands of Industrial Internet of Things (IIoT) applications. However, the presence of autonomous mobile robots (AMRs) in industrial environments introduces additional complexity due to mobility-induced interference, fading conditions, and intermittent line-of-sight (LoS). This paper develops an integrated MEC-assisted IIoT framework that incorporates AMRs, small-cell base stations (SBSs), macro-cell base stations (MBSs), and edge caching capabilities. Within this framework, we investigate how different association mechanisms influence downlink energy efficiency. Three representative association strategies are considered: a distance-oriented approach, a cache-aware method that prioritizes local content availability, and a cloud-centric policy. To quantify their performance, we derive semi-analytical expressions that capture the effects of industrial propagation factors and network heterogeneity. The numerical evaluation reveals that cache-driven association markedly enhances energy efficiency compared to strategies that do not exploit edge caching, underscoring the importance of cache-centric designs in future IIoT deployments.