From Active Names to π-calculus Rewriting Rules

Abstract

The problem of checking equivalences for π-agents is not trivial and has been widely studied in the last decade. Syntactic and semantic approaches can be taken to formally verify π-calculus equivalences. The syntactic approach rests mainly on structural congruence. On the other hand, the semantic checking methods can verify wider equivalences but cannot check infinitary π-agents. Bisimilar agents have the same set of active names. This result and a technique to check bisimulation considering active names is presented in [U. Montanari and M. Pistore. Checking bisimilarity for finitary π-calculus. In Insup Lee and Scott A. Smolka, editors, Proceedings of CONCUR '95, volume 962 of LNCS, pages 42–56. Springer, 1995]. There, agents active names are calculated from their corresponding Labelled Transition Systems (LTS) and, because of this, cannot be directly applied to rewriting systems. In [Ana C.V. de Melo. A study on the potential active names of π-agents. ENTCS (Electronic Notes in Theoretical Computer Science) 95 (C) (apr 2004) 269–286], a syntactic characterisation of active names for π-agents was presented. Here, new rewriting rules are presented (based on the syntactic characterisation of active names) to identify and discard useless code of π-expressions for a class of expressions including composition. With these new rules, π-expressions are better reduced (more useless code is discarded) enriching the equivalence classes of agents.