Research On Load Balancing And QoE Based SDN Path Selection Algorithm

In order to cope with the pressure of processing performance brought by rigid traditional network structure and simplified functions.Based on the novel network architecture — —SDN(Software Defined Network),this paper optimizes the network path through joint consideration,acquisition and weighting calculation of the dual impact factors affecting link state,increases the consideration of server-level factors affecting network load balancing,and achieves network load balancing more effectively.On this basis,focusing on the actual QoE(Quality of Experience)that the end users pay more attention to nowadays,we discuss the transmission path selection of the specific multimedia traffic(video stream)in the network.The innovative contributions of this paper include the following three aspects:(1)A path selection algorithm for SDN networks based on double influence factors is proposed.In this algorithm,the corresponding data acquisition strategies are constructed for link delay and link available bandwidth,which affect the real-time status of links.The optimal real-time link is determined by assigning different weight coefficients to different impact factors.The simulation results show that the proposed SDN network path selection algorithm based on double influence factors has higher average bandwidth utilization,lower transmission delay,more uniform load distribution of destination host and the number of times selected compared with the traditional Dynamic Load Balancing algorithm based on single hop optimal greedy strategy.(2)Based on the content of the previous innovation,in the ant colony optimization-based SDN network load balancing algorithm proposed in this innovation,additional consideration is given to the server-level factors affecting network load balancing,and the server real-time state is quantified by the self-defined performance index——SBD(Server Busy Degree),to achieve the real-time state of the server.Fixed real-time optimal server.At the same time,considering the path level,ant colony algorithm,a meta-heuristic algorithm widely used in the field of path search,is used to determine the path to the selected server.The simulation results show that the proposed load balancing algorithm for SDN network based on ant colony optimization has higher data transmission rate and lower packet loss rate than the Shortest Path First Random algorithm.(3)Based on the above two innovations,in the proposed SDN network path selection algorithm based on QoE measurement results,additional considerations are added to characterize the impact of QoE on the network path selection,and research is carried out.In view of the distortion characteristics of video streams in the network transmission process,the corresponding QoE evaluation strategy is selected to complete the effective measurement and representation of QoE quantification results;ant colony algorithm and QoE measurement results are combined to determine the path that meets the requirements of the proposed algorithm.The simulation results show that the proposed SDN network path selection algorithm based on QoE measurement results has larger QoE measurement results(objective evaluation scores)than the Shortest Path First algorithm.