HIQCO: A Hierarchical Optimization Method for Computation Offloading and Resource Optimization in Multi-Cell Mobile-Edge Computing Systems

Zhiyong Li,Chen Du,Shaomiao Chen

doi:10.1109/access.2020.2977988

Abstract

Mobile-Edge Computing (MEC) addresses the shortcomings of mobile devices in terms of computing performance and energy efficiency by offloading computation tasks to the edge of the network. In this paper, a multi-cell MEC system is considered, where each small cell connects to a common MEC server, and the problem of minimizing system overhead in the MEC system is solved by jointly optimizing the computation offloading decisions, communication and computational resources. This problem is a mixed integer nonlinear programming problem (MINLP). Owning to the combined nature and complexity of this problem, we adopt a method to transform the original problem into two sub-problems: (i) joint optimization computation offloading decision and channel assignment problem (JO-CODCA) and (ii) joint optimization transmission power and computational resource allocation problem (JO-TPCRA). For the transformed sub-problems, we propose a hierarchical optimization method (HIQCO) that combines immune algorithm (IA), quasi-convex optimization and convex optimization techniques. Experimental results show that the proposed HIQCO give better performance comparing with the other compared algorithms in terms of system overhead.

Highlights

With the rapid development of the Internet and the popularity of mobile devices, a large number of mobile applications have emerged, such as online games, virtual reality and mobile medical [1]–[3]
Our research focuses on building a system model of multi-user, multi-cell, and a single mobile-edge computing (MEC) server
We study the computation offloading management system model in the multi-cell MEC system

Summary

Introduction

With the rapid development of the Internet and the popularity of mobile devices, a large number of mobile applications have emerged, such as online games, virtual reality and mobile medical [1]–[3]. These emerging applications require a large amount of computing and storage resources and are sensitive to the time delay of task processing [4]. To tackle these challenges, a novel computational paradigm of mobile-edge computing (MEC) is proposed. Offloading and resource allocation schemes become an important issue in MEC [8], [9]

Methods

Results

Conclusion