Energy-Efficient Trajectory Design and Computing Offloading in UAV-Aided IoT Networks

In this paper, we propose to minimize the total energy expenditure in unmanned aerial vehicle (UAV) assisted mobile edge computing (MEC) internet of things (loT) networks by employing the non-orthogonal multiple access (NOMA). Considering the obstacles of an urban city environment, a composite optimization model incorporating both the flight trajectory and computational offloading is initially formulated. To address this non-convex problem, the original problem is converted into two independent subproblems, and then an energy-efficient trajectory design and computing offloading algorithm is developed by employing the successive convex approximation (SCA) and quadratic approximation approaches. Moreover, comprehensive simulation results verify that the proposed algorithm achieves a decrease in weighted energy expenditure by 14.8% in comparison to the algorithms employing a fixed trajectory, fixed transmit power, and the orthogonal multiple access (OMA).

I. Introduction

The unmanned aerial vehicle (UAV) has been used exten-sively in internet of things (IoT) networks for data gathering, processing and calculating due to its features of simple implementation and low cost [1]–[3]. With the advent of mobile edge computing (MEC), the smart mobile devices (SMDs) can transfer their computational workload to servers housed on the UAVs. However, the UAV and SMDs are usually energy limited, which poses a great challenge on the joint optimization of flight trajectory and computing offloading to burst out the application value of UAV s in IoT networks [4].