Abstract
Mobile Edge Computing (MEC) is an innovative solution designed to address key challenges in mobile cloud computing, including latency, limited capacity, and resource constraints. The primary objective of MEC is to enable dynamic scheduling and efficient resource allocation with minimal cost. This paper proposes a three-tier system architecture comprising mobile devices, edge computing nodes, and traditional cloud infrastructure. It introduces two methods for task offloading and scheduling. For task allocation on mobile devices, the system leverages the Greedy Auto-Scaling Offloading algorithm, which prioritizes high-energy-consuming tasks to enhance energy efficiency. On the edge computing layer, a dynamic scheduling approach based on fuzzy logic is presented, which ranks and allocates tasks according to two specific criteria. Numerical evaluations demonstrate that, compared to existing alternatives, the proposed method significantly reduces task waiting time, latency, and overall system load, while maintaining system balance with minimal resource consumption. Moreover, the proposed system achieves up to ~ 64% reduction in battery consumption in our simulated environment compared with local execution. The results also indicate that over 93% of tasks are successfully executed within the edge environment.