Research On Product Derivation Techniques In Software Product Lines

In the recent years, with the widely application of information, commu-nication and Internet, software industry has deeply impact the developmentof other social industries. Therefore, how to generate the software meetingthe customers‘requirements quickly has become a very important issue af-fecting the development of the companies. In the earlier, it is a bit difficult togenerate the software meeting the customers‘requirements. However, due tothe proposal of software product lines, it is now possible to generate thesoftware products quickly in specific domains.The key part in the process of software derivation in software productlines is selecting the proper features from feature model generated in domainengineering. Then, we use the selected features to compose the new softwareproduct. It is a time consuming process. Especially for the situation with re-source constraint, it is proved to be an NP-hard problem. This paper is focuson how to generate the software product meeting the customers‘requirementson the basis of software product lines.This paper presents a new method for the existing problems in the tradi-tional products derived technology in software product lines. The first step ofthe method is transforming the feature model to a directed graph and the fea-ture selection problem turn into routing problem. Then apply the ant colonyoptimization algorithms to solve the routing problem. We design a series ofexperimental simulation to evaluate the performance of the method. Weevaluate our approach by comparing it to another two important approxima-tion technique FCF+M-HEU and GAFES. The experimental results show that our approach performs6%worse than FCF+M-HEU with reducing muchrunning time. Meanwhile, it performs10%better than GAFES with takingmore time.