|
Queueing theory is a useful discipline for modeling computer systems, and mean-value analysis (MVA) is one of its most powerful tools. We present two network models for computer systems, together with algorithms to solve them. We show how certain features of Mathematica, such as functional and rule-based programming and the ability to set up a database with symbolic indices, make the implementation of the exact MVA algorithm for closed, multiclass queuing networks much easier than it is in traditional programming techniques.
|
|