Proving system properties by means of trigger-graph and Petri nets

Abstract

The formal description of hardware protocols allows system engineers to check these protocols for some important properties, such as completeness and consistency. In this paper, a methodology called trigger graph is introduced. It uses a notation that is similar to timing diagrams, but it strictly avoids their disadvantage of being ambiguous. In addition, the exact mathematical notation of the trigger graph gives the possibility to transform it into Petri nets. This gives the benefit to use methods developed for Petri nets also for the check of hardware protocols, for example to check them for security conditions. The trigger graph methodology is illustrated with three realistic examples: the control of traffic lights, the control of a railway-barrier, and the read-protocol for a DRAM.