Abstract

1
(No Transcript)
2
Abstract

• Self-similar processes based on fractal point
  processes (FPP) provide natural and attractive
  network traffic models.
• FPP models provide useful in evaluating and
  predicting the queuing performance of various
  types of fractal traffic sources.

3
Introduction

• Various stochastic models and techniques have
  beeen proposed for modeling the distinct
  statistical nature of self-similar network
  traffic.
• EX Brown motion ,Markovian models

4
Introduction

• Four FPP
• 1.fractal renewal process(FRP)
• 2.superposition of several fractal renewal
  processes (Sup-FRP)
• 3.fractal-shot noise driven poisson process
  (FSNDP)
• 4.fractal binomial noise driven process (FBNDP)

5

• Provide two methods which yield parsimonious
  flexible FPP
• One based on renewal processes and the other on
  doubly stochatic poisson processes.

6
FPP

• Fractals are objects whitch process a form of
  self-similarity parts of the shole can be made to
  fit to the whole in some way by scaling.

7
Seocond order statistcal measures

• PSDpower spectrum density
• CRcoincidence rate
• IDCindex of disperation for counts(fano
  factor)
• COVcount-based covariance dunction.

8
(No Transcript)
9

• ? ENT/T is expected rate

10
2.3Fractal nature of FPPs
11
(No Transcript)
12
(No Transcript)
13
3.Two FPP construction method

• 3.1 Renewal point process method
• 3.2 Doubly stochastic poison process

14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)