Machine Learning Based Traffic Prediction and Congestion Control Algorithms in Software Defined Networks

Author:
Publisher website
Google Scholar
Abstract
With the rise of Software Defined Networking (SDN) and the tremendous success of Machine Learning (ML) methods in classification, prediction, and control tasks, finding new traffic engineering techniques to adaptively and dynamically manage or route traffic in networks, ensuring service quality, and improving user experience quality has become a hot research topic in network research. SDN is a new type of network architecture with enormous growth potential. Its core idea is to separate forwarding and control, and concentrate intelligence on the controller for centralized control. This paper proposes an ML based traffic prediction and congestion control algorithm to address the problem of uneven load on the control plane of SDN architecture due to the complexity and variability of network traffic. Historical traffic data is collected from the SDN controller and feature extraction is performed. The extracted features are used to train the ML model and predict future traffic. Based on predicted traffic information and perceived congestion, the controller can dynamically adjust network parameters to avoid or alleviate congestion. The experimental results show that the algorithm proposed in this paper can achieve more efficient network traffic management and congestion control in SDN, thereby improving network performance and user experience.

This publication has 12 references indexed in Scilit: