Abstraction of parallel uniform processes with data

Abstract

In practice, distributed systems are quite often composed by an arbitrarily large but finite number of processes that execute a similar program. Abstract interpretation is an effective technique to fight state explosion problems. In this paper, we propose a general framework for abstracting parallel composition of uniform processes with data, in the setting of a process algebraic language /spl mu/CRL We illustrate the feasibility of this technique by proposing two instances of the general framework and applying them to the verification of two systems.