Research on Traffic Optimization Scheme of SDN Network Based on ME-HMM. Issue 4 (October 2020)

• Record Type: Journal Article
• Authors: Yang, Yanchun; Sun, Hongfeng
• Is Part Of: Journal of physics. Volume 1624:Issue 4(2020)
• Publication Date: 2020-10
• Subjects: Software-Defined Networking -- Maximum entropy -- Hidden Markov -- Optimization; Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1624/4/042052
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: In this paper, a traffic optimization scheme based on Maximum Entropy- Hidden Markova Model (ME-HMM) in Software-Defined Networking (SDN) networks is proposed in order to resolve the load imbalance in SDN architecture. First, the data flow generated by the host in the SDN network was monitored, and denoising and analyzing would be carried out with the maximum entropy (ME) model to predict the network traffic for a period of time in the future, then the controller would be deployed. Next, the traffic flow was evaluated with Hidden Markova Model (HMM) and the deployment was optimized in the SDN control plane. With the feedback information, the scheme could optimize SDN network traffic and promote load balancing.