Analysis and Solution of Generalized Stochastic Petri Nets: a Data Parallel CM-5 Approach

摘要

Generalized Stochastic Petri Nets (GSPN) represent an assessed formal methodology for performance analysis.of cuncurrrent systems. The applicability of this methodology is severely limited by potential state space explosion phenomenon. In this paper we describe a CM-5 based data parallel tool dedicated to the complete numerical solution of GSPN models. The main effectiveness of this approach is the significant extension of state space cardinalities that can be generated and solved in reasonable amounts of execution times.