Multi-resource fair allocation with bandwidth requirement compression in the cloud–edge system. (January 2023)

• Record Type: Journal Article
• Authors: Li, Xingxing; Li, Weidong; Zhang, Xuejie
• Is Part Of: Computers & electrical engineering. Volume 105(2023)
• Publication Date: 2023-01
• Subjects: Cloud–edge collaboration computing system -- Multi-resource allocation -- Bandwidth requirement compression -- Fair allocation -- Dominant resource fairness; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.108510
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: For the multi-resource fair allocation problem in cloud–edge collaborative computing systems, we propose a mechanism called dominant resource fairness in the cloud–edge collaborative computing system (DRF-CE). DRF-CE enables users to deploy tasks to cloud and edge servers and considers the particularity of cloud server bandwidth resources. By introducing bandwidth demand compression, DRF-CE allows users to have two different bandwidth resource demands. DRF-CE satisfies envy-freeness, Pareto efficiency, strategyproofness and the weak sharing incentive property. To implement DRF-CE, we design an algorithm and evaluate its performance via simulations driven by a concrete example and Google cluster traces. Simulation results prove that DRF-CE is superior to traditional approaches in terms of users' dominant share, resource utilization and the number of tasks performed by each user. Furthermore, the simulation results show that considering bandwidth demand compression can improve user efficiency by approximately 40% and the central processing unit (CPU) utilization by more than 25%. Graphical abstract: Highlights: Fair multi-resource allocation is realized in the cloud–edge collaborative computing system. Bandwidth demand compression is considered when tasks are processed on different server levels. Users are allowed to have two different bandwidth resource demands. Envy-freeness, Pareto efficiency, strategyproofness and the weak sharing incentive property are satisfied.