A Design and Implementation of Energy-Aware Resilience Architecture for Mobile Edge Cloud

초록

In In edge-cloud environments, mobile nodes face significant challenges due to their mobility and the distributed nature of the environment. The unstable communication links between mobile nodes and the cloud often lead to frequent disruptions in connectivity, posing obstacles to seamless operation and service delivery. Effective energy management strategies are crucial to address these challenges and ensure the long-term viability of mobile nodes. In this paper, we propose an architecture for energy-aware resilience in edge-cloud environments for a standalone mobile node in an edge-cloud environment that can operate seamlessly in the connection disruption from the cloud. Our architecture leverages machine learning-based energy consumption prediction techniques to forecast energy consumption patterns while considering dynamic network conditions. In addition, we propose a threshold-based control policy for autonomous node resilience, enabling mobile nodes to adaptively adjust their operations in response to fluctuating energy levels and network conditions of edge environments. Through proactive energy management strategies, such as workload autoscaling with energy awareness, we aim to minimize energy consumption and maximize node survival time, particularly under constrained conditions. Experimental evaluations demonstrate the efficiency of our proposed approach in extending node longevity and ensuring reliable operation in dynamic and resource-constrained edge-cloud environments.

키워드