On the Distributed Implementation of Petri Nets

By an implementation of a Petri net (place/ transition system) we understand an algorithm which carries out the occurrence of the transitions in accordance with the available tokens, i.e. an algorithm which executes the token game. In this paper fundamental studies are made, how such an implementation can be carried out in a distributed manner. In order to compare different distributed software implementations we start out from an explicitly defined basis, called the process paradigm. Quality criteria are specified for implementations which are based on this paradigm. Subsequently a variety of implementations of finite place/transition systems is presented and investigated with respect to the quality criteria. Proposals for implementations of Petri nets found in the literature are transferred to our framework and discussed. The programs are formulated in Occam. © 1988 De Gruyter Oldenbourg. All rights reserved.