Research On Behavior-Driven Development Based On Problem Frames

In recent years,with the development of information technology,we will see increasing use of the Internet of things(IOT),big data analytics systems and cloud computing systems,and the appearance of Cyber-Physical Systems,such as intelligent traffic management systems,intelligent road lamp management systems and intelligent dam control systems.The requirements analyses of CPS projects,not only need to consider the functional requirements of software,but also need to consider the physical characteristics of hardware equipment,and the electrical characteristics of engineering equipment.So the requirements analysis of CPS projects becomes more and more important.The requirements analysis in the early phase of software development will directly affect the whole project's progress and the quality of software.Therefore,in the process of software development.the requirements provider and the software project team should be paid more attention and allocated adequate personnel and funds to complete the tasks in the requirements analysis phase.Research on Problem Frames(PF)is one of the important hot spots in requirements engineering.During the last several decades,PF has been widely recognized as a major approach to software requirements analysis.In this approach software requirements are defined as a set of interactions between the software systems and real-world problem domains,software requirements are treated as part of a problem,and the development of software is a process of solving problems.In PF,a problem is regarded as a software requirement in its real-world context,whose structure is described by a problem diagram and the problem is solved by using problem transformations based on cause-and-effect relationships.Although many research results have been produced in Problem Frames,how to transform requirements models(i.e.,problem diagrams)into software design artifacts and achieve implementations smoothly remains an open problem.Behavior-driven-development(BDD)is one of the best practices of extreme programming,which is an outstanding achievement in the development process of software engineering.In order to avoid the problems caused by inconsistent expression,this approach encourages both the demand provider and the software project team to use the "common language" to describe the behavior of the system.In order to ensure the quality of the software,this approach recommends that testing codes be generated as early as in the requirements engineering phase before implementing the software.Requirements analysis is a key factor in a software project.However,BDD lacks a requirements analysis modeling method in the requirement analysis phase.Therefore,it is of great significance to provide a requirement analysis modeling method for BDD.This dissertation proposes a software development method which combines Problem Frames and Behavior-Driven-Development(PFBDD)for Cyber-Physical Systems.It uses Problem Frames approach for requirements modeling,which can automatically generate a textual description(User Scenarios)from requirements models.User scenarios are used as a bridge between Problem Frames and Behavior-Driven-Development,which links them closely.The method proposes a way of transforming requirements models into software design artifacts and achieving implementations smoothly,which will play an important role in driving Problem Frames into further practical applications.In PFBDD,a formal language named Alloy is used to describe requirements models in the requirements modeling phase,in order to build the corresponding relationship between the requirements model and its formal semantics,while using the Alloy Analyzer to verify the requirement model,so as to ensure the correctness of the requirements model.In order to ensure consistencies among the modelling,design and implementation codes in the PFBDD development process,this dissertation also provides a requirements tracking process,which establishes a tracking chain to ensure traceability in the development process.In addition,a computer-aided tool for design and implementation is developed to support PFBDD,which implements a completeness check algorithm for problem diagrams and an algorithm to support the automatic conversion of a problem diagram into user scene texts.Finally,a case study is presented to demonstrate how our method is applied to a queuing problem in a vehicle management system.At the same time,the PFBDD tool is also used in that case study,and the result demonstrated the feasibility and practicality of the PFBDD approach.