One algorithm for time-domain analysis of semi-Markov processes

This paper describes one algorithm of the time-domain analysis of a large class of stochastic processes, called semi-Markov processes, which are characterized by the Acyclic Signal Flow Graphs, (ASFG). This method is based on the minimization of the number of the convolutions. In that way the time domain analysis of ASFG can be efficiently achieved. MATLAB program was made as a support of this algorithm. The input data are the matrices of the ASFG and the output is the time gain of the ASFG.