Latency optimization of task offloading in NOMA‐MEC systems

Abstract

This paper investigates low‐latency offloading strategy in a non‐orthogonal multiple access aided mobile edge computing (NOMA‐MEC) system consisting of K edge servers, one mobile user and one cloud server. An intelligent edge server selection strategy (IESSS) based on Markov decision process (MDP) is proposed to select an edge server, in order to reduce the task completion latency of this system. When an edge server is selected by the proposed IESSS, a joint optimization problem of power allocation and task scheduling factors is formulated to minimize the task completion latency of the hybrid NOMA‐MEC system. To solve the formulated non‐convex optimization problem with coupled variables, a low‐complexity adaptive power‐task resource allocation iterative (APTRAI) algorithm is proposed. Simulation results demonstrate the advantages of the proposed IESSS and verify the convergence and time complexity of the proposed APTRAI algorithm.

Delay-Optimal Multi-Satellite Collaborative Computation Offloading Supported by OISL in LEO Satellite Network

“…Low Earth orbit (LEO) satellite networks have received widespread attention in recent years [1], due to its advantages of wider coverage area and lower propagation delay [2]. With the considerable advancement of satellite tech-nologies on manufacturing [3], optical inter satellite link (OISL) and on-board computing capability [4], orbital edge computing holds great promise in addressing the computation demand of resource-limited users [5] located in remote areas without proximal servers or in disasterstricken areas that lose terrestrial network connections [6].…”