Research Article
A Self-adaptive PSO-Based Dynamic Scheduling Method on Hierarchical Cloud Computing
@INPROCEEDINGS{10.1007/978-3-030-48513-9_7, author={Shunmei Meng and Weijia Huang and Xiaolong Xu and Qianmu Li and Wanchun Dou and Bowen Liu}, title={A Self-adaptive PSO-Based Dynamic Scheduling Method on Hierarchical Cloud Computing}, proceedings={Cloud Computing, Smart Grid and Innovative Frontiers in Telecommunications. 9th EAI International Conference, CloudComp 2019, and 4th EAI International Conference, SmartGIFT 2019, Beijing, China, December 4-5, 2019, and December 21-22, 2019}, proceedings_a={CLOUDCOMP}, year={2020}, month={6}, keywords={Dynamic scheduling PSO Mobile edge computing Cloud Workflow}, doi={10.1007/978-3-030-48513-9_7} }- Shunmei Meng
Weijia Huang
Xiaolong Xu
Qianmu Li
Wanchun Dou
Bowen Liu
Year: 2020
A Self-adaptive PSO-Based Dynamic Scheduling Method on Hierarchical Cloud Computing
CLOUDCOMP
Springer
DOI: 10.1007/978-3-030-48513-9_7
Abstract
Edge computing has been envisioned as an emerging and prospective computing paradigm for its advantage of low latency, which uses local resources. However, the edge resources are usually limited and could not meet end-users’ diversified requirement. Cloud computing paradigm could provide scalable and centralized resources with high computational capabilities, but it has latency issues. Thus it is suggested to combine both computing paradigms together to improve the performance of mobile applications. In this paper, we propose a self-adaptive dynamic scheduling approach based on hierarchical heterogeneous clouds. Our scheduling mechanism considers not only schedule planning but also dynamic scheduling on heterogeneous clouds. Firstly, a self-adaptive scheduling mechanism based on a meta-heuristic optimization algorithm, PSO (Particle Swarm Optimization), is presented for schedule planning. Then a dynamic scheduling mechanism on dynamic partial workflow model is proposed for dynamic optimization during the execution. Finally, external experiments compared with other methods are conducted to demonstrate the effectiveness of our proposal.
