| At present,the deployment of transportation power plan continues to be promoted in China.And the development of railway is also changing from high-speed large-scale development to high-quality.In order to improve the railway transportation efficiency and traffic safety furtherly,National Railway Administration of the People’s Republic of China puts forward new requirements for the time synchronization accuracy of railway transportation.Railway time synchronization network,as one of the supporting networks for railway communication,plays the role of time synchronization in each subsystem link.As the transmission network and data communication network carrying railway time synchronization,the transmission traffic is complex and diverse.So it is difficult for the uplink and downlink to achieve delay symmetry in the process of time synchronization.This problem reduces the time synchronization accuracy based on Network Time Protocol(NTP).Thus,the efficiency of railway transportation will be reduced,even the safety of running will be endangered.However,there are few studies on improving the accuracy of railway time synchronization network,and the existing research mainly focuses on the accuracy of time synchronization by improving the structure of railway time synchronization network and ensuring the reliable operation of protocols.There is still a gap in the study of link delay symmetry.How to improve the synchronization accuracy while ensuring the efficiency and performance of railway transmission network is an important topic.Therefore,based on the calculation and analysis of queuing delay,this thesis proposes a queuing delay compensation scheme based on Stochastic Flow Queue-Control Delay(SFQ-CoDel)algorithm.The scheme can reduce the queuing delay while ensuring the overall performance of the network.Firstly,aiming at solving the problem of delay asymmetry,this thesis conducts delay analysis,determines several delays that affect data transmission delay in the transmission link,and then puts forward a clock synchronization model for railway time synchronization network.Considering the difficulty in obtaining offset parameters accurately and the clock model for railway time synchronization network,this thesis improves the clock offset calculation in Real Time Calculation Queue Delay(RTCQD)algorithm,and proposes a new algorithm based on Enhanced Real Time Calculation Queue Delay(ERTCQD).The time complexity and effect of the algorithm are analyzed and verified.It shows that the calculation error of the improved algorithm for queuing delay is lower,which is only 0.22 ms.To comprehensively analyze the expected compensation effect of queuing delay,combined with the railway time synchronization model mentioned before,the simulation analysis is carried out in OPNET network simulation platform.The results show that eliminating the queuing delay can not only reduce the asymmetry of the two-way delay,but also improve the time synchronization accuracy.Secondly,based on the calculation and analysis of queue delay,this thesis proposes a queue delay compensation method based on SFQ-CoDel algorithm.The algorithm uses the advantages of AQM in the field of congestion control to reduce the queuing delay.Through the analysis of the traffic in the transmission network derives the traffic type,and the traffic type is used as a simulation scenario in NS2 platform to simulate and validate the algorithm.The dumbbell topology model is used in the simulation,and two kinds of simulation environments are set up for high and low loads respectively.Three performance indexes under two states are obtained and analyzed respectively.The results show that the packet loss rate,throughput and queue length of SFQ-CoDel algorithm are better than the other two algorithms,so it proves the feasibility of the scheme.Finally,this thesis analyzes the adjustability of SFQ-CoDel algorithm,and proposes a queuing delay compensation scheme.The scheme provides a theoretical reference for improving the accuracy of railway time synchronization network and the performance of transmission network. |