Overload Analysis for Markovian Models
A new procedure for computing stationary
probabilities for an overloaded Markovian model is proposed in
terms of the rotated Markov chain.
There are two advantages to use this procedure:
i) This procedure allows us to approximate an overloaded finite model
using a stable infinite Markov chain. This will make
the study easier when the infinite
model has a simpler solution.
ii) Numerically, this procedure often significantly reduces the number of
computations and the requirement of computer memory.
By using different examples,
we specifically demonstrate
the process of implementing this rotating procedure.
Back to my
home page.