(Source section: network traffic management
File format in compressed package: e-book
Attachment source: Internet
Running platform: Windows platform
Whether it has been verified by me: Yes
Nature of accessories: Free
Multi scale analysis of TCP traffic
Absrtact: Based on wavelet technology, the global scaling and local scaling properties of multi group measured data of LAN and WAN traffic are studied. discuss
Fractal Gaussian noise model and multifractal wavelet model are proposed. In the study of global scaling properties, a class of Hurst parameter estimators is discussed
Compared with fractal Gaussian noise model, it is found that network traffic on large time scale can be approximated by fractal Gaussian noise. In the local scaling property
In the research, the local scaling index and the corresponding structure function are analyzed, and the actual data are compared with the multifractal wavelet model
In fact, neither fractal Gaussian noise nor multifractal wavelet model can well approximate network traffic, and the division of large and small time scales is in a typical round trip
Around time.
Keywords: wavelet; Scaling properties; Fractal Gaussian noise; Multifractal; Network traffic modeling
This document contains the following attachments:)