New convergence results on functional iteration techniques for the numerical solution of M/G/1 type Markov chains (Q1374752)

The aim of the paper is to perform an accurate analysis and to give a complete theoretical explanation on the convergence and experimental behaviour of functional iteration techniques used in the process of computing the minimal nonnegative solution \(G\) of the matrix equation \(X= \sum^\infty_{i= 0} X^iA_i\). This equation provides the numerical solutions of M/G/1 type Markov chains, where \(A_i\) are nonnegative \(k\times k\) matrices such that the matrix \(\sum^n_{i=0} A_i\) is a column stochastic one. The paper introduces the general class of functional iteration methods and estimates the rate of convergence of the sequence generated in this class, for the case \(X_0=0\). Finally, functional iteration methods are investigated for the case where the starting point \(X_0\) is a column stochastic matrix.