Balanced Cloud Edge Resource Allocation Based on Conflict Conditions

Leilei Zhu,Xiaolong Song,Hongwei Yang,Zhengqi Bai,Jiahui Feng,Li Li,Dan Liu

doi:10.1109/access.2020.3032428

Leilei Zhu, Xiaolong Song + Show 5 more

Open Access

PDF Available

https://doi.org/10.1109/access.2020.3032428

Copy DOI

Export

Save

Cite

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

Affiliation: Changchun University of Science and Technology

Abstract
Highlights/Summary
Full-Text PDF
Similar Papers

Abstract

Listen

Under a multiscenario environment with frequent bursts of data in the edge cloud, the resource allocation in the edge cloud will affect the stability of its nodes. To address this problem, a balanced virtual resource allocation model based on conflict conditions is proposed. Based on a thorough study of the similarity between task attributes and resources used by the host, the concept of a task backlog is implemented to achieve a preliminary balanced allocation of tasks; thus, a conflict condition based on the remaining resources of the physical and virtual machines is proposed. Further, a matrix of phased conflict coefficients is built to establish a balanced virtual machine allocation model. The results of experiments comparing the performance of the proposed model with that of other existing models indicate that the proposed model can reduce the virtual machine scheduling time by up to 8.33%, save up to 6.25% of host energy consumption, and improve the algorithm efficiency by 20.47% compared with the other algorithms. To avoid the local optimal problem caused by dynamic virtual machine migration, an improved ant colony algorithm is combined with the above model, and concepts of a pheromone volatility factor and suppression factor are implemented to optimize the pheromone measurement function and ensure that the virtual machine migration path is globally optimal. Overall, the model reduces the conflict rate of resources on the physical machine and can maintain stable operation under CPU usage fluctuation, thus realizing a balanced allocation of node resources.

Highlights

Edge clouds are cloud computing platforms built on edge infrastructure based on the core of cloud computing technology and edge computing capabilities
To address the abovementioned issues, this study proposes the SimCMA model, which considers the effect of physical machine thresholds on resource allocation
We ignore the network bandwidth under the edge cloud; we focus on the relationship between pheromones and VM scheduling

Summary

Introduction

Edge clouds are cloud computing platforms built on edge infrastructure based on the core of cloud computing technology and edge computing capabilities. These platforms may have a distributed cloud architecture (i.e., peer model) [1]. With the increase in the number of Internet of things devices, the traditional way of processing delivered resources by the central cloud has been unable to satisfy the requirements of tasks generated by edge devices. An established edge cloud model can accept the task of edge device delivery for the Internet of things so that the edge device can obtain the nearby resources dynamically and quickly to meet the

Objectives

Results

Conclusion