Online Computation Offloading in NOMA-Based Multi-Access Edge Computing: A Deep Reinforcement Learning Approach

Maurice Nduwayezu,Quoc-Viet Pham,Won-Joo Hwang

doi:10.1109/access.2020.2997925

Maurice Nduwayezu, Quoc-Viet Pham + Show 1 more

Open Access

https://doi.org/10.1109/access.2020.2997925

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2020
Citations: 40	License type: CC BY 4.0

Affiliation: Pusan National University, Inje University

Abstract

One of the missions of fifth generation (5G) wireless networks is to provide massive connectivity of the fast growing number of Internet of Things (IoT) devices. To satisfy this mission, non-orthogonal multiple access (NOMA) has been recognized as a promising solution for 5G networks to significantly improve the network capacity. Considered as a booster of IoT devices, and in parallel with the development of NOMA techniques, multi-access edge computing (MEC) is also becoming one of the key emerging technologies for 5G networks. In this paper, with an objective of maximizing the computation rate of an MEC system, we investigate the computation offloading and subcarrier allocation problem in Multi-carrier (MC) NOMA based MEC systems and address it using Deep Reinforcement Learning for Online Computation Offloading (DRLOCO-MNM) algorithm. In particular, the DRLOCO-MNM helps each of the user equipments (UEs) decides between local and remote computation modes, and also assigns the appropriate subcarrier to the UEs in the case of remote computation mode. The DRLOCO-MNM algorithm is especially advantageous over the other machine learning techniques applied on NOMA because it does not require labeled data for training or a complete definition of the channel environment. The DRLOCO-MNM also does avoid the complexity found in many optimization algorithms used to solve channel allocation in existing NOMA related studies. Numerical simulations and comparison with other algorithms show that our proposed module and its algorithm considerably improve the computation rates of MEC systems.

Highlights

In 5G and beyond, user equipments (UEs) are expected to run compute-intensive, latency-sensitive, and energy-hungry applications
With an objective of maximizing the system computation rate, we propose an Online Computation Offloading using deep reinforcement learning (DRL) algorithm to solve the problem of offloading decision and subcarrier allocation in Multi-carrier non-orthogonal multiple access (NOMA)-enabled multi-access edge computing (MEC) systems (DRLOCO-MNM)
4) We evaluate the performance of our DRLOCO-MNM algorithm by comparing it with orthogonal multiple access (OMA) (TDMA)-based algorithm, optimal algorithm, edge computing and local computing algorithms

Summary

INTRODUCTION

In 5G and beyond, user equipments (UEs) are expected to run compute-intensive, latency-sensitive, and energy-hungry applications. From the studies in [32], [33], our work considers multi-carrier NOMA enabled MEC systems and proposes a DRL algorithm, by which both offloading decisions and subcarrier assignment are optimized. With an objective of maximizing the system computation rate (in terms of bits processed over a given time duration), we propose an Online Computation Offloading using DRL algorithm to solve the problem of offloading decision and subcarrier allocation in Multi-carrier NOMA-enabled MEC systems (DRLOCO-MNM). Our proposed DRLOCO-MNM algorithm helps many UEs decide their offloading modes by considering few actions, and assigns subcarriers without complex optimization problems It can serve more UEs by solving their computation modes (i.e., local computing at the UE or edge execution at the MEC server) and subcarrier allocation by considering only two binary actions 0 or 1 for each UE.

COMPUTATIONAL MODEL

PROBLEM FORMULATION

23: Uniformly sample a batch of data set

PROPORTION OF UEs IN LOCAL COMPUTATION

CONCLUSION