Design And Implementation Of A Circular-compromise Negotiations Framework That Based On Game Theory And Used For Eliminating The Requirements Inconsistency

In software engineering, a consistent set of system requirements is the basis for softwareproject success. Requirements change is normal happen in today’s software projects, changesin requirements may cause a blurred defined set of requirements, a collection elementredundancy, a collection incompletely and a consistency set and so on, these issues are themain impact factors of the requirements’ quality and software’s quality. Once the requirementsspecification created, new demand requests may result in inconsistent demand systemproblems. Generally speaking, the fusion of old and new set of requirements should improvethe system’s problems, and eliminate the inconsistencies, these thoughts are important andurgent.The theoretical ideas of game theory is inconsistent conflict transplanted into softwareengineering, the software needs change request can be considered as caused by changes indemand with the original demand for the development of a conflict of interest between: Bothsides hope to own set of requirements as the system demand collection, and hope does notproduce changes in demand inconsistencies. Therefore, the use of game theory strategy andideological inconsistency can handle changing requirements, get consistent results of thenegotiations.The main work of this paper are the following:1. Based on Game Theory, using a analytical methods for software engineering needs ofcustomers actual needs and game theory formalized system for symbolic language in formalsign language system after using the negotiation rules matrix game, pure strategy Nashequilibrium, Nash equilibrium mixed strategy of negotiation strategies and selection processcomputing, control changes in demand due to satisfactory negotiations and negotiating thebest results.2. Based on compromise and concessions ideas, build compromise and concessionsalgorithm for inconsistencies that may arise negotiations, the use of reasonable concessionrules gradually narrow the problem domain, by eliminating inconsistencies generated everyconcession, to eliminate the inconsistent results of the negotiations. 3. Based on the outcome of negotiations and compromise matrix game algorithm in thecontent, establishing a framework based on game theory to negotiate a compromise cyclethinking, the architecture through the cycle of negotiations will eliminate inconsistencies existin the system, get consistent requirements specifications.4. Based on an actual case on Software Engineering (a residential area access controlsystems), first given their natural language requirements specification, followed by convertingit into a logical language requirements specification, using a matrix proposed bargaining gameagain rules and compromise and concessions inconsistencies may occur between negotiatingalgorithm processing requirements, the final result will be restored to natural languageprocessing requirements specifications. In this case demonstrates the feasibility of thisarchitecture.5. Design and implement an application-system to eliminate the inconsistent. First, giventhe needs of several groups of customers’ needs as the basic data entered into the system.Second, by architecture for computing proposed, given the results of the negotiations and theoutput matrix game results of the last system. Finally, design of the architecture of theproposed experiment with two typical architecture already exist were compared, this paperanalyzes the architecture of the validity and superiority.Based on the changing needs of existing management methods and commonmanagement rules introduced for the analysis and discussion of demand in the currentrequirements specification inconsistencies, proposed a compromise based on game theorycirculating negotiating framework. The architecture introduces a technology-relatednegotiations selection strategy game theory as the rules and ideological compromise andconcessions as the basic algorithm to eliminate inconsistencies that can make up fordeficiencies in the traditional selection policy issues such as flexibility and accuracy, improveelimination system requirements inconsistent. Finally, the results of an instance of a softwareengineering associated comparative analysis proved the correctness and feasibility of thisarchitecture.