The protocol engineering plays an important role in computer networks. Designing a protocol is a challenging activity because of the complexity of the rules defining the interaction between the communicating entities. At this point, formal approaches have been proposed: Analysis and Synthesis. The synthesis approach has the advantage of avoiding errors a priori and ensuring some desirable properties during the protocol design process. In this paper, we propose a relational protocol synthesis method. The method outcome is specifications of all protocol entities serving the different service access points, which are represented in R-net diagrams.

Recent research results in software engineering have shown that the process of formalizing and automating software specification activity is moving towards knowledge-based specification systems. A modeling approach and a knowledge-based system that assist in the requirement acquisition, development, and maintenance of such software system specifications are presented.>

The requirement specification activity is one of the most error-prone and expensive in the software development life cycle. It is also a complex activity because of the fuzzy and informal nature of the user requirements, the lack of automated tools to support it, and the requirements evolution. A deductive approach is presented. It deals with some fundamental AI issues regarding the knowledge acquisition process-in particular, the way informal requirement descriptions evolve into formal ones. To show how the approach can be applied, a fragment of a patient monitoring system is specified in the top-down decomposition technique. The approach is implemented in a requirement specification tool, SMES (software modeling expert system).>