Research And Application On Algorithm Of Traffic Control And Congestion Management In QoS

Traditional transmission network technology can not meet customer demand along with increasing and diversification of customer business,packet transport network(PTN)technology came into being,and QoS has been developed faster and faster.When overload or congestion in the network channel,the technology of QoS can not only provide different levels of service to different business,but also partly reduce loss of business and delay of network to guarantee the quality of service for business customers.The processing of QoS includes: flow classification,flow control,traffic shaping,congestion avoidance,queue scheduling,and so on.Data will be discarded in the process of traffic regulation and congestion avoidance of the data packet processing,which determines the data packets discarded in the network transmission directly,and also influences the delay and jitter in the network.So it has very important practical significance and application value to study the algorithm of traffic control and congestion avoidance in QoS in the two processes.First of all,this thesis describes the technical features of PTN technology and the network layer,analyzes the QoS processing flow of PTN,and designs the system overall plan of QoS.Then,we mainly research the algorithm of traffic control and congestion avoidance in QoS,and analyze the token bucket algorithm in traffic regulation from theory and practice.We also debug and analyze parameter setting of token bucket algorithm,and optimize the parameter in the actual product application.In order to improve the utilization rate of network resource flow in the process of supervision proposed a new token sharing algorithm,we test and analyze the token bucket algorithm and the token sharing algorithm,aiming at the shortcomings in the original algorithm.We test and analyze advantages and disadvantages about the algorithm of tail drop,random early detect and weighted random early detect in congestion avoid.According to the demand of the actual network environment,the combination of the tail drop algorithm and the WRED algorithm is put forward to solve the problem of network congestion avoidance.At last,flow regulation and congestion avoidance are verified from the point of the product development,the results show that the algorithm can achieve the QoS function and meet the requirements of the relevant performance indicators.At the same time,we test the token sharing algorithm in traffic control and the combination of the WRED algorithm and the congestion avoidance algorithm,and analyze the results.