Font Size: a A A

Sampling Measurement Research For Traffic Of Campus Network Export Based On Mask Matching

Posted on:2006-05-26Degree:MasterType:Thesis
Country:ChinaCandidate:M S TanFull Text:PDF
GTID:2168360155962558Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
As the rapid development of network technology and the emergence of the high-speed Ethernet, it is increasingly difficult to measure the full-grouping of the Network Traffic directly, restricted by the factors such as the memory capacity and processing speed etc. Thus sampling measurement technology becomes the most important research in the high-speed network traffic measurement field. This paper studies the sampling measurement technology based on the mask matching in the background of campus network.Aiming at the randomicity of IP message identification field in campus network export, this paper applies the conception of entropy in the information theory to define the random metric. The paper also focuses on the effects of bit flow length and position of the identification field to the random metric, and the effects of the specific message(the IP message has the length of 40 bytes with identification field value of 0)to the random metric of the bit flow. The result of the experiment shows that with the increase of the length ,the random metric of bit flow tends to reduce gradually. In respect of the same length of the bit flow , right priority random metric is universally the highest, and the left priority one is universally the lowest with others between the two. And special messages have relatively great influence on the random metric of bit flow of identification field. If these messages are processed in another way, not only the random metric of the bit flow has obvious improvement but also with small variation range, and the length distribution of message which is sampled before and after with protocol distribution become more and more consistent.This paper provides the architecture of the sampling measurement system based on the mask matching for the purpose to design and achieve the sampling measurement system scientially. This architecture divides the system into three levels which achieved by java including message collecting, sampling, saving. This paper mainly analyzes execution time of program about the message sampling and saving level when it adopts difference temporary-memory ways, and compares the longest delayed time by three different temporary-memory ways. The paper gives the choice of measurement point and the design project of measurement environment aiming at the problem of execution and application of the mask match sampling measurement in the campus network, and execute the right priority match sampling measurement about special mask and...
Keywords/Search Tags:Mask Matching, Network Traffic, Sampling Measurement, Random Metric, Campus Network
PDF Full Text Request
Related items