Probability, statistics, and queueing theory with computer science applications
Probability, statistics, and queueing theory with computer science applications
Web server workload characterization: the search for invariants
Proceedings of the 1996 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Internet Web servers: workload characterization and performance implications
IEEE/ACM Transactions on Networking (TON)
Generating representative Web workloads for network and server performance evaluation
SIGMETRICS '98/PERFORMANCE '98 Proceedings of the 1998 ACM SIGMETRICS joint international conference on Measurement and modeling of computer systems
A large-scale study of file-system contents
SIGMETRICS '99 Proceedings of the 1999 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
A compound model for TCP connection arrivals for LAN and WAN applications
Computer Networks: The International Journal of Computer and Telecommunications Networking - Special issue: Advances in modeling and engineering of Longe-Range dependent traffic
On the relationship between file sizes, transport protocols, and self-similar network traffic
ICNP '96 Proceedings of the 1996 International Conference on Network Protocols (ICNP '96)
Self-configuring network traffic generation
Proceedings of the 4th ACM SIGCOMM conference on Internet measurement
Computer Networks: The International Journal of Computer and Telecommunications Networking - Special issue: Long range dependent trafic
Tmix: a tool for generating realistic TCP application workloads in ns-2
ACM SIGCOMM Computer Communication Review
Lognormal and Pareto distributions in the Internet
Computer Communications
Hi-index | 0.00 |
A novel unique probability distribution, which has a lognormal body and either light or heavy tail, has been fitted to various empirical data sets of Web file sizes. The optimal parameters of this distribution have been determined by the maximum likelihood estimation combined with the optimization algorithm minimizing a goodness-of-fit metric specially adopted to provide the best fit to the upper tail. The mirror transformation of the processed original data set with respect to the median has been proposed to improve the fit. The obtained results question the common opinion that the probability distribution of Web file sizes is heavy-tailed. The lognormal fits are given for comparison.