Research On Transmission Optimization And Traffic Assignment Mechanism Based On Intelligent SDN

With the emergence of various intelligent terminals and applications,and the rapid growth of Internet users,the network traffic shows explosive growth.It has put forward new demands on the network.The emergence of Software-Defined Networking(SDN)and Network Function Virtualization(NFV)technology will reshape the entire network architecture.This thesis attempts to bring artificial intelligence technology into the new network architecture.We research on transmission optimization and traffic assignment,and hope to create a more intelligent and efficient network.First of all,this thesis proposes intelligent SDN system architecture.By introducing intelligent center,it provides richer control functions and intelligent decision making capabilities for SDN.The efficiency of information sharing between multi domain SDN controllers is also improved.Then,this thesis presents a transmission optimization system based on intelligent SDN,it optimizes the performance of data transmission primarily through intelligent path selection,important node analysis and traffic forecasting.Intelligent Path selection is to find the optimal path for the end to end data transmission according to the network state,and ensure the quality of data transmission.Important node analysis enhances the stability of the entire transmission system by strengthening the important nodes in the network.Traffic forecasting refers to the real-time prediction of node traffic according to the historical traffic data in the network,and realizes dynamic scheduling to avoid congestion.Also,under the architecture of intelligent SDN,this thesis optimizes the traffic assignment in Service Function Chaining(SFC)of Network Function Virtualization.This thesis proposes an algorithm combining Artificial Neural Networks(ANNs)and reinforcement learning.It implements the assignment of user traffic intelligently into different SFC,and reduces the average waiting time of the system.