GMB-ECC: Guided Measuring and Benchmarking of the Edge Cloud Continuum

Modern edge–cloud deployments integrate devices, edge servers, private networks, and cloud platforms into heterogeneous, multi-layer systems. Each layer exposes distinct metrics, measurement granularities, and time scales, which complicates system-wide reasoning about energy use. Coordinated cross-layer optimization can reduce energy consumption, but assessing efficiency is difficult because metrics are incompatible and measurements are noisy. Practitioners must navigate efficiency trade-offs when selecting among alternatives, yet often rely on isolated measurements and system-specific heuristics. This paper introduces GMB-ECC (Guided Measuring and Benchmarking of the Edge–Cloud Continuum). GMB-ECC is a framework for precision-aware energy-efficiency benchmarking in heterogeneous edge–cloud systems. It models deployments as directed acyclic graphs that enforce conservation when aggregating energy and work metrics. The framework derives normalized efficiency measures with uncertainty bounds and ranks optimization opportunities by gap confidence, energy shares, and remediation costs. We evaluate GMB-ECC in a synthetic intra-logistics scenario and a live industrial deployment. The industrial deployment includes autonomous mobile robots, an edge server, and a private 5G network. Our framework-guided changes reduced total service energy consumption by 12% relative to an energy-unaware baseline while meeting latency and safety constraints. These results indicate that precision-aware benchmarking can support cross-layer energy optimization in heterogeneous edge–cloud environments.