Dear Colleagues, Many research questions in networking are described by probability theory formulated with measure theory. Recently we find that measure-theoretic probability appears to be inconsistent. The inconsistency has caused undesirable consequences in applications including network traffic modeling. Our findings are presented in a technical report.
The title and abstract are as follows. Reconsidering Network Traffic Modeling We question current techniques ofnetwork traffic modeling. Measure-theoretic probability is the foundation of current traffic models. However, due to a fundamental flaw of measure-theoretic probability, current traffic models cannotproperly characterize the growth trend of network traffic. By demonstrating theinconsistency of measure-theoretic probability implied by unbounded random variables, we justify theabove claim with a rigorous analysis, and discuss the implications of our analysis to traffic modeling and some other related issues, such as performance evaluation of traffic control. The technical report is available at http://www.eee.hku.hk/research/doc/tr/TR2012006_Reconsidering_Network_Traffic_Modeling.pdf Your comments will be very much appreciated. Best regards, Guang-Liang Li and Victor O. K. Li _______________________________________________ IEEE Communications Society Tech. Committee on Computer Communications (TCCC) - for discussions on computer networking and communication. [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/tccc
