Research On Video Streaming Load Balancing Based On SDN Networks

In recent years,with the rapid development of Internet,new types of networking services and applications,such as video streams,have emerged.Streaming media applications such as 4K videos,VR videos require predictable,steady network resources with Minimal delay variation and few packet lose.Software Defined Networking(SDN)paradigm is an emerging networking paradigm that separates the network control plane from data forwarding plane,which promotes solutions of effectively manage and direct traffic.There is a novel frame using SDN to solute network congestion and to meet end-to-end users' Quality of Service(QoS)requirements.This paper presents a simple solution of dynamic Quality of Service(QoS)which based on buffer memory in client.SDN controller gets the Network information,such as Topology,Delay,used Bandwidth,etc.Using SDN controller and Openflow,network administrator can get global view and achieve global optimization.More importantly,this solution also collects the buffer memory in client.The buffer memory in client inform how long the video can be play,and a better evaluation of network states.It's the buffer that decide when the network should be optimized.And the network information decide how the network should be optimized.This paper used Simulated-Annealing(SA)algorithm with backfire to optimized the flows.Using this algorithm,solution getting a better speed while finding a forwarding solution.With SAA,controller balance network resources in real time to avoid congestion.This paper uses network simulation software Mininet and RYU as controller to build and set up the environment.Using a simple website provide 4k VOD server.This experimental results show that buffer memory in client can predict and inform the network states.This Solution can provide better QoS than only focus on the network states.