Modeling Internet Traffic Packet Length Using Probdistid: A Case Study




Abstract:
In this study, we apply the ProbDistID tool, a user-friendly tool based on nonlinear regression, designed for fitting probability distributions and estimating their parameters, to model internet traffic packet length using a real-world internet traffic dataset. The tool requires no a priori knowledge of input data, making it suitable for real-time fitting recognition and for data mining tasks. Our primary objectives in this case study are to identify distributions that offer the best fit for internet traffic datasets. We utilized our tool to fit and estimate parameters for eight cumulative density functions (CDFs). The fitting results are presented using utilized several model selection methods and goodness-of-fit tests to determine the most appropriate distribution model. The case study indicate that the Generalized Extreme Value (GEV) and Pareto distributions provide the most accurate fit. Our findings are presented graphically and in tabular form, demonstrating the effectiveness of ProbDistID and its potential applicability across various fields, including data mining tasks.

CITATION:

IEEE format

D. Miljković, S. Ilić, B. Jakšić, P. Milić, S. Pitulić, “Modeling Internet Traffic Packet Length Using Probdistid: A Case Study,” in Sinteza 2023 - International Scientific Conference on Information Technology, Computer Science, and Data Science, Belgrade, Singidunum University, Serbia, 2023, pp. 172-177. doi:10.15308/Sinteza-2023-172-177

APA format

Miljković, D., Ilić, S., Jakšić, B., Milić, P., Pitulić, S. (2023). Modeling Internet Traffic Packet Length Using Probdistid: A Case Study. Paper presented at Sinteza 2023 - International Scientific Conference on Information Technology, Computer Science, and Data Science. doi:10.15308/Sinteza-2023-172-177

BibTeX format
Download

RefWorks Tagged format
Download