SUPERPOSED STOCHASTIC AUTOMATA - A CLASS OF STOCHASTIC PETRI NETS WITH PARALLEL SOLUTION AND DISTRIBUTED STATE-SPACE

Stochastic processes generated by the composition of stochastic automata can be analyzed efficiently (both in terms of space and time) with a technique based on Kronecker (tensor) algebra for matrices, as shown by Plateau. The technique is applied in this paper to the analysis of a class of Stochastic Petri nets (SPN) that is named Superposed Stochastic Automata (SSA) an it is used to solve nets of hundreds of thousands states. The evaluation of the steady state probability distribution for this class of models was implemented with sequential as well as parallel algorithms. Although SSA are a rather restricted subclass of SPN, the extension of the analysis methodology to a more general setting appears to be feasible.