Research And Implementation Of IP Network Delay Measurement And Estimation Techniques Research And Implementation

As the representative IP network plays an increasingly important role in people’s lives. The Internet technology is constantly developing, the network size is constantly expanding, the degree of isomerization of the network is increasing that makes our network environment more complex. Network delay is very significant for network troubleshooting, traffic monitoring, and many aspects of the network topology design, so the IP network delay measurements, estimates and analysis has an important significance.This paper firstly used a variable step size discrete delay model to reduce the delay quantization interval for the problem of the calculation complexity higher of the link delay estimation algorithm PLE problem, thereby it reduced the computational complexity of the link delay estimation algorithm PLE, and on the basis of NS-2simulation experiments we verified the accuracy and efficiency of the improved algorithm; secondly we described the design and implementation of the subsystem of a distributed network quality monitoring and route optimization system that is the distributed network quality monitoring system. The subsystem is the source and basis of a large number of data, delay data is also measured and used for analysis and calculation of the route optimization system, thereby completing the network quality monitoring and optimization features; Finally, by analyzing and modeling the IP network delay data, we discovered that delay distributions approximately meets Weibull model and found the general rule delay distribution, and used Matlab simulation software to fit round-trip delay data and verified the rationality and effectiveness of the delay distribution model.Through this paper, the improved PLE algorithm of link delay estimation reduced the computational complexity, and the "distributed network quality monitoring system" achieved a better network delay measurement function, so they can be widely applied to future network delay measurement and estimation.