Traffic Schedule Technology Research And Implementation On Multiple QoS Constraints

With the rapid development of Internet, Internet users are increasing and they demand for network service quality also gradually improve.Network resources is inadequate or even the occurrence of network congestion, resulting in network flow not to be distributed evenly.How to reasonable use of valuable network resources, achieve load balancing has the vital significance. Network traffic scheduling is using an effective can enhance network performance and improve the availability of the method on the basis of the existing network,which is in the server can accept processing capacity to distribution of network traffic, and in the process of dispatching multiple QoS constrained minimum. Therefore, it has a good research value and practical significance to design a stable and reliable traffic scheduling algorithm with multiple QoS constraints.This thesis, a multi-objective optimization method is used to solve the traffic scheduling problem. we first research the traditional traffic scheduling related technologies, Then analysis of the characteristics of multiple QoS constraints, and on the basis of the network traffic data collection and analysis, we describ the influence of multiple QoS constraints for traffic scheduling.we convert flow scheduling problem into a multi-objective optimization problem and the traffic was established based on multi-objective optimization scheduling. According to the established mathematical model, we established the multi-objective optimization function, including the time delay function, flow equilibrium function, hop function.Based on the above we combined with Pareto sorting mechanism and the based artificial bee colony algorithm, a multi-objective optimization colony algorithm, then we proposed a multi-objective bee colony optimization algorithm,we used it for traffic schedule on multiple QoS constraints to Implement multiple traffic scheduling program. Finally, determine the best traffic scheduling scheme according to the actual situation, implement network traffic scheduling efficiently on multi QoS constraint.In this paper, we test it in simulation and compare it with the multi-objective genetic algorithm, the experimental results show that this algorithm scheme has better effect.Finally,through designing virtual traffic scheduling system to show traffic scheduling program intuitively.