ON THE ANALYSIS OF FILE TRANSMISSION DURATION DISTRIBUTION OVER THE WEB

摘要

The rapid growth of the World Wide Web as a medium for information dissemination has made it important to understand its characteristics, in particular the distributions of its file sizes and files transmission duration. It is well known that the distribution of files transmission duration in the Web is heavy tailed. Also one more important problem related to the file size and transmission direction that is scheduling connection on a busy web server. This research attempts to understand the reasons for this phenomenon by measurement study and analysis of the distribution of the transmission duration for each file separately. We made use of several sanitized world data in addition to enormous pool of data we recorded from our testing network. In this research we prove that the transmission duration distribution of the same file from the same server to the same client in the Web is Pareto and therefore heavy tailed. We present results showing that for a specific client/server pair, transmission delay is not significantly affected by file size. Namely, all files transmitted from the same server to the same client have a very similar transmission duration distribution, which surprisingly, stays similar even for files with different order of sizes. In addition to that, special attention is given to investigate the dependence of the transmission duration distribution on server load. Also we find also that the file size and the transfer time are increasingly correlated as the file size increases. In addition to the absolute difficulty in predicating transmission time from the file size only.