Research On Computational Offloading Algorithms With Reliability Guarantee In Multi-access Edge Computing

Multi-Access Edge Computing(MEC)is an emerging concept in network technology that aims to push computational resources closer to users.In recent years,a new trend in computing has been happening that more and more cloud functions move towards the edge of the network.With tens of billions of edge devices deployed at the edge of the network in the future,it allows the network edge to harvest a large amount of free computational resources and storage space that can generate sufficient execution ability to handle computationally intensive and latency-critical tasks on user devices.MEC is a promising paradigm.It not only provides services similar to cloud computing,but also can handle high-complexity and latency-sensitive applications on user equipment(UE)by computational offloading.However,the reliability requirements are rarely considered in the current MEC research,which is also an important factor to ensure the quality of service(Qo S).For this reason,this paper designs two energy-saving offloading algorithms to satisfy the reliability and latency constraints of applications.The main contribution includes:1: First,this paper analyzed the current research of multi-access edge computing technology and introduced in detail the development of multi-access edge computing technology based on reliability requirements.And then this paper given an introduction to the offloading technology in multi-access edge computing,finally,established an optimization model to minimize the energy consumption of the UE under the premise of satisfy the reliability and latency constraints.2: In order to solve this NP hard problem,this paper proposes the first energy-saving offloading algorithm,namely the fast heuristic algorithm.First,it divides the entire application into multiple sub-tasks based on Directed Acyclic Graph(DAG).These sub-tasks can be executed on the local UE or MEC server.Then the overall reliability and latency constraints of the application are decomposed into multiple constraints for each subtask,and finally,it can get a solution that minimizes UE energy consumption among the solutions that satisfy the reliability and latency constraints.3: This paper proposes the second energy-saving offloading algorithm,the greedy algorithm.First,it divides the entire application into multiple DAG-based subtasks.These subtasks can be executed locally on the UE or on the MEC server.Then,according to the greedy strategy,after the current subtask is executed,if the constraints conditions are satisfied,then the execution continue,otherwise the process returns to the previous task.Until all subtasks are completed.