Analysis Of Network Time Synchronization Characteristics Based On Measurement Data

Time synchronization is an important prerequisite in many networks.For different network application environments,there are different requirements for the accuracy of time synchronization.For example,time management in daily life requires accuracy to be maintained at the second level,while network measurement requires accuracy to be maintained at the millisecond level,and the network-based real-time control environment of the Industrial Internet and smart grid has higher accuracy requirements.With the rapid development of science and technology,while enjoying the convenience brought by science and technology,people have also put forward higher requirements for the synchronization accuracy of communication networks.Many different network clock synchronization technologies have been born,such as the Network Time Protocol(NTP),the Global Positioning System(GPS),and the Precision Time Protocol(PTP).NTP and GPS technology can meet most of the daily network timing accuracy requirements.For network environments with higher timing accuracy requirements,PTP is the choice not to be taken seriously.With the development of technology,the requirements for timing accuracy in the Industrial Internet and Smart Grid have further increased,especially the time sensitive network(Time Sensitive Networking,TSN)..With the improvement of computer hardware capabilities,computational analysis of large-scale data has become one of the most popular analysis methods at the moment.This paper uses the rich hardware equipment provided by the laboratory to collect and test the timing performance of the PTP protocol in different networks and conduct a comparative analysis;analyzes the system state equation and data distribution of the data,and builds a neural network to complete the data fitting.Innovatively use data to analyze related PTP protocols and propose new ideas for the subsequent use of data to improve the performance of PTP.The specific work is as follows:(1)Investigated the related research on the improvement of PTP timing performance,and related factors affecting the accuracy of PTP timing.Through related research,the goals of this analysis were determined,and an analysis plan for the timing accuracy of PTP was proposed.(2)Using the rich hardware network equipment and OSA high-precision time server in the laboratory,the timing performance test of the PTP protocol was performed on different current networks,and the timing accuracy of different networks was compared and analyzed.Then,the impact of the number of hops and load on the timing accuracy of PTP is analyzed and evaluated.(3)Based on the collected experimental data,construct a system state equation of the data,analyze the distribution of the data,and use the Long Short-Term Memory(LSTM)distribution curve of the data over time in deep learning.The fitting was performed,and the experiment proved that it has a good fitting effect,which provides a new idea for the subsequent improvement of PTP timing accuracy using the data.