Joint Task Offloading and Resource Allocation for Energy-Constrained Mobile Edge Computing
Hongbo Jiang,Xingxia Dai,Zhu Xiao,Arun Iyengar,Arun K. Iyengar
DOI: https://doi.org/10.1109/tmc.2022.3150432
IF: 6.075
2022-01-01
IEEE Transactions on Mobile Computing
Abstract:We consider the problem of task offloading and resource allocation in mobile edge computing (MEC). To maintain satisfactory quality of experience (QoE) of end-users, mobile devices (MDs) may offload their tasks to edge servers based on the allocated computation (e.g., CPU/GPU cycles and storage) and wireless resources (e.g., bandwidth). However, these resources could not be effectively utilized unless an encouraging resource allocation scheme can be proposed. What’s worse, task offloading incurs additional MEC energy consumption, which inevitably violate the long-term MEC energy budget. Considering these two challenges, we propose an online joint offloading and resource allocation (JORA) framework under the long-term MEC energy constraint, aiming at guaranteeing the end-users’ QoE. To achieve this, we leverage Lyapunov optimization to exploit the optimality of the long-term QoE maximization problem. By constructing an energy deficit queue to guide energy consumption, the problem can be solved in a real-time manner. On this basis, we propose online JORA methods in both centralized and distributed manners. Furthermore, we prove that our proposed methods enable the achievement of the close-to-optimal performance while satisfying the long-term MEC energy constraint. In addition, we conduct extensive simulations and the results show superiority in performance over other methods.
computer science, information systems,telecommunications