modelling of real network traffic by phase-type distribution · andriy panchenko dresden university...
TRANSCRIPT
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 1
M&SM&SM&S
Modelling of Real Network Traffic by Phase-Type distribution
Andriy PanchenkoDresden University of Technology
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 2
M&SM&SM&S
Outline
1. Problem overview2. Fitting algorithm3. Experiments
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 3
M&SM&SM&S
Traffic Problem overview
Need of a traffic modelsimple but adequateapplicable to further performance analysis
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 4
M&SM&SM&S
Failure of Poisson process
- Interarrivals are not exponentially distributed- Heavy tailed distributions
- Presence of huge values that are generated with small probability but have a great impact on the lower moments of pdf.
Problem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 5
M&SM&SM&S
Traffic modeling goals
• Goal: to develop a traffic model• simple but adequate• applicable to further analysis (possibly by means of
existing tools and techniques)
• Approaches• Separate modeling of each transmitting process and their further
aggregation (simulation)
• Statistical description of traffic processes(analytical model)
• …
Problem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 6
M&SM&SM&S
Modeling compromiseR
eal t
raffi
c tra
ces
Exist
ing
tool
s and
an
alys
is te
chni
ques
Model
FGn,Wavelets
New techniques
?
Problem overview
Markovianprocesses
Model parameterisation
?
Poisson
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 7
M&SM&SM&S
Phase-Type (PH) distribution
i
h
k
j
D0i,j(rates)
N+1
Absorbing state
• π is the stationary distribution (π Q=0) • τ the distribution after an arrival
complete definition of PH-distribution
• CTMC with N states described by a generator matrix:• D0 includes local transitions rates and negative diagonal elements• d1 includes transition rates accompanied by an arrival
d1i
Problem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 8
M&SM&SM&S
Properties of PH-distribution
• distribution function
• probability density function
• k-th moment
Problem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 9
M&SM&SM&S
• computing of CTMC transient measures
• Randomisation
where
• Randomisation rate
• Poisson probability
Randomisation techniqueProblem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 10
M&SM&SM&S
General fitting task
• Adjust the free parameters of PH-distribution
• Maximize the density function over all values
Problem overview
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 11
M&SM&SM&S
• t – sample; O – model parameters• EM iterative algorithm for maximum likelihood parameter
estimation if data are incomplete or hidden.• Problem: find maximum likelihood estimate OML
• Principle: sequential mapping such that • E-step: evaluation of conditional expectation• M-step: maximization
• Differentials w.r.t. model parameters O• Generalized EM
EM-algorithm (Expectation-Maximization)
Fitting algorithm
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 12
M&SM&SM&S
Exploited features
• Aggregated trace (approx.)
• Fixed initial distribution
• Reduced Likelihood
• Exploit the randomization technique
Fitting algorithm
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 13
M&SM&SM&S
Empirical pdf
• Represent trace t as set of tuples
• 1: Identical intervals- Not enough intervals- Heavy tails
• 2: Log-scaled intervals
Fitting algorithm
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 14
M&SM&SM&S
Steps of our EM-algorithm
1. start with some PH-distribution2. compute the likelihood of the PH-distribution
according to the each element in the trace3. compute the new transition probabilities in the
matrices for the given likelihood4. check for convergence and stop
or repeat steps 2-4.
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 15
M&SM&SM&S
EM-steps
• Forward likelihood
• Backward likelihood
Fitting algorithm
k
x
y
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 16
M&SM&SM&S
EM-steps
• Collect weighted values of likelihood
• Normalize
Fitting algorithm
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 17
M&SM&SM&S
Algorithm framework
Input: Trace, Initial distribution, Randomization rateBuild aggregated traceEncode matrices
REPEATCalculate forward / backward likelihoodsFOR all intervals
Calculate likelihood matrices END forCollect and normalize likelihood matrices in
UNTIL converges
Decode matrices
Fitting algorithm
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 18
M&SM&SM&S
Influence of initial distributionResults
uniform(1,0,…,0)random
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 19
M&SM&SM&S
Influence of matrix formResults
dense\\upper triang
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 20
M&SM&SM&S
Fitting of heavy-tailed distributionsResults
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 21
M&SM&SM&S
Conclusion
• good capture of main part as well as tail behaviour of empirical distribution
• a larger number of phases yields a better likelihood value
• best approximation was achieved with • initial vector (1,0,…,0) • matrix form: upper triangular or with main and first
upper diagonals• h.t.d. properties are captured with relatively
compact models with 5 states
-
27-28.Juli.2004 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" 22
M&SM&SM&S
Thank You!