A Two-Stage Hybrid Multi-Objective Optimization Evolutionary Algorithm for Computing Offloading in Sustainable Edge Computing

Edge computing is an effective complementary technology to cloud computing, allowing end devices to offload tasks onto edge base stations (BSs) to satisfy the quality of experience of consumers. Due to the limitation of storage and computing resources, a single BS cannot satisfy the heavy computing...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on consumer electronics 2024-02, Vol.70 (1), p.735-746
Hauptverfasser: Li, Lingjie, Qiu, Qijie, Xiao, Zhijiao, Lin, Qiuzhen, Gu, Jiongjiong, Ming, Zhong
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Edge computing is an effective complementary technology to cloud computing, allowing end devices to offload tasks onto edge base stations (BSs) to satisfy the quality of experience of consumers. Due to the limitation of storage and computing resources, a single BS cannot satisfy the heavy computing tasks. In this regard, multi-BS collaboration is an effective way to alleviate this issue. Moreover, service caching and cloud-edge collaboration computing also show attractive advantages in handling the surging data traffic. However, to the best of our knowledge, there is rarely work that consider all of the aforementioned scenarios simultaneously. To fill this research gap, this paper comprehensively considers the computing offloading problem in sustainable edge computing based on the above scenarios. Specifically, the computing offloading problem is first modeled as a multi-objective optimization problem with the purpose of minimizing the delay and energy consumption. Then, a two-stage hybrid multi-objective optimization evolutionary algorithm, called TH-MOEA, is designed to address the above formulated problem, which uses a novel competitive swarm optimizer to accelerate convergence in the early evolutionary stage and adopts a diversity-enhanced immune algorithm to improve diversity in the later evolutionary stage. Simulation results show that TH-MOEA outperforms several state-of-the-art peer methods.
ISSN:0098-3063
1558-4127
DOI:10.1109/TCE.2024.3376930