Task offloading based on two types of Edge-Edge collaboration in mobile edge computing

Task offloading based on two types of Edge-Edge collaboration in mobile edge computing

연구 분야: Networking

학회: Computing

Task offloading research in Mobile edge computing (MEC) has experienced significant growth over the past years. Users commonly utilize MEC servers for task offloading in order to achieve Quality-of-Service standards, including low delay and low energy consumption. Due to the restricted computational capability of a single MEC server, task offloading between different MEC servers with constrained resources has become increasingly popular for the Edge-Edge collaboration mechanism. Currently, research on Edge-Edge collaboration is divided into two types: vertical Edge-Edge collaboration and horizontal Edge-Edge collaboration. Nevertheless, there has been limited focus on merging these two types of offloading techniques. To address the aforementioned issue, this paper proposes a multi-layer MEC-MEC model that employs Edge-Edge Collaboration in both Vertical and Horizontal (EEdge-CoVeHo) directions for task offloading. Next, a multi-objective optimization function is constructed to balance the system delay and the system energy consumption. Subsequently, we propose a hybrid heuristic algorithm called Particle swarm Simulated annealing-based Ant colony optimization. Last but not least, a simulation environment to verify our proposed task offloading scheme outperforms in many aspects when compared to various baseline methods.