Software Design and Testing Using Petri Nets : A Case Study Using a Distributed Simulation Software System *

This paper presents the application of Petri Nets in software design and testing using a distributed simulation application for naval radar interference resolution as a case study. Good software systems design promotes the “minimal” exertion of control over the different subsystem components; nevertheless, the knowledge of “critical” states of operation within one subsystem can orient other subsystems to react intelligently. Hence, a methodology for determining all the important “logical areas” of a subsystem is most crucial. The approach used here is based on a technique called Minimal Transition Cover Sets, or MTCS, which is a refinement of the minimal transition invariants of the corresponding Petri Net model of the system [1]. It allows for selective information sharing; one of the most important attributes of good object-oriented design. This paper explores this approach to aid the system designer build some high level knowledge about the different subsystem components. Using this high-level knowledge, the designer can then choose an appropriate design from a group of designs and thus can be in a position to qualitatively compare and contrast two or more design in terms of information sharing capabilities [2]. A further enhancement of this MTCS approach involves its use in building software test cases. It allows for the development of the necessary test cases in the verification of software system modules.