Computation Efficiency Maximization in Multi-UAV-Enabled Mobile Edge Computing Systems Based on 3D Deployment Optimization

Unmanned aerial vehicles (UAVs) have been widely devoted to mobile edge computing (MEC) systems that have limited resources to provide high-quality computing and communication services for Internet of Things (IoT) terminals. Energy-efficient computation and resource allocation are key issues for the sustainable operation of the above-mentioned systems. The 3D deployment optimization of multi-UAVs is also crucial to maximizing the system's computation efficiency. In this paper, we discuss a multi-UAV-enabled MEC system. To maximize the computation efficiency of the terminal system, we consider jointly optimizing the terminal's CPU frequency, transmission power, offloading correlation decision, and the 3D position and beamwidth of the UAV. Since the original problem is a mixed-integer nonlinear programming (MINLP) problem with a fractional structure, which is difficult to solve directly. Based on Dinkelbach's method, convex optimization theory, and greedy strategy, we simplify the mathematical model and propose a four-stage alternating iterative computation efficiency maximization algorithm(FICEM) to solve the problem. The simulation results indicate that the algorithm converges fast in various network scenarios and that its computation efficiency is better than that of the baseline algorithm. In addition, the simulation results also manifest the effect of different network parameters on the computation efficiency of the terminal system.

I. Introduction

In Recent years, the vigorous development of Internet of Things (IoT) technology has led to the emergence of many intelligent applications such as intelligent agriculture, intelligent industry, and intelligent transportation, which makes a large number of daily stuff such as automobiles, household appliances, mobile phones, and traffic lights enter the network for data exchange [1]. This expansion will lead to a tremendous increase in the network scale and data volume [2], which far exceeds the existing network capacity. To support a large number of IoT terminals and to process a large amount of data in time, mobile edge computing (MEC) [3] was born, and it supports the delay-sensitive and computation-intensive applications by providing ubiquitous computing resources near mobile user devices [4], [5], [6].