MAM-Net : a petri-net based approach to software process modeling, analysis, and management

Abstract

In the field of software development, there have been enormous effort to improve the quality of software products and the productivity of software development. As one of the results from these efforts, a new development approach called "process-driven software development" has been introduced. The main advantage of process-driven software development is the ability to explicitly guide and staff development processes. The underlying assumption of this approach is the existence of fine-grained formal process model. Unfortunately, the traditional process modeling and analysis techniques do not support such formal modeling of processes. This lack of fulfillment raised a new issue, formal techniques to software process modeling. Recently, several formal approaches to software process modeling have been introduced. Each of them has its own advantages and disadvantages in terms of representing various aspects of software processes. Our study focuses on developing a new Petri-net based approach, namely MAM-Net(Modeling, Analysis and Management-Net) that can model organizational properties as well as other properties in a uniform way. As a formal software process modeling approach, MAM-Net provides formal modeling constructs for representing process activities, process artifacts, and process control flows. MAM-Net also supports the modeling of process agents and their control flows in process models in order to reflect organizational properties. In analysis, not only the standard Petri-net analysis techniques are supported, but also new additional analysis techniques have been developed in order to provide helpful information for project planning and management. With these modeling and analysis capabilities, MAM-Net can be used in software process modeling, analysis, and management. This thesis presents the formal definitions and process modeling mechanisms of MAM-Net. In addition, the supporting analysis capabilities are described.