Class Snc_env_map
For a Markovian arrival process (D0,D1) counting N(0,t) unit-work jobs,
E[exp(theta*N(0,t))] = pi*expm((D0+D1*exp(theta))*t)*1. With lstar
the eigenvalue of maximal real part of A(theta)=D0+D1*e^theta and
v > 0 its right Perron eigenvector, bounding 1 <= v/min(v)
entrywise gives
rho(theta) = lstar/theta, sigma(theta) = log(max(v)/min(v))/theta,
the standard exponential-form envelope of a Markov-modulated source. The
burst term is what the modulating chain contributes: 0 for a one-phase MAP,
where this reproduces Snc_env_poisson exactly, and positive for an
MMPP.
THE PERRON PAIR IS COMPUTED BY POWER ITERATION ON THE SHIFTED MATRIX
A+cI, not by a general eigensolver. A(theta) is essentially
nonnegative, so the shift makes it nonnegative with a positive diagonal,
hence primitive whenever the MAP is irreducible, and the iteration converges
to the pair the bound needs without asking a general solver which of its
eigenvectors is the positive one. The MATLAB reference uses eig and
agrees to machine precision.
Port of matlab/src/api/snc/snc_env_map.m. Reference: C.-S. Chang, Performance Guarantees in Communication Networks, Springer 2000, Ch. 7.
-
Method Summary
Modifier and TypeMethodDescriptionstatic SncEnvelopestatic double[]snc_env_map(Matrix D0, Matrix D1, double theta)
-
Method Details
-
snc_env_map
- Parameters:
D0- hidden-transition generator block of the MAPD1- arrival-transition block of the MAPtheta- Chernoff parameter, theta > 0- Returns:
- {sigma, rho}
-
of
- Parameters:
D0- hidden-transition generator blockD1- arrival-transition block- Returns:
- the envelope as a function of theta
-