基于可靠性数据差异的软件可靠性模型比较与分析  

Comparison of Software Reliability Models Based on Differences in Reliability Data

作　　者：臧思聪 皮德常[1] 

机构地区：[1]南京航空航天大学计算机科学与技术学院,南京210016

出　　处：《小型微型计算机系统》2016年第2期252-258,共7页Journal of Chinese Computer Systems

基　　金：国家自然科学基金民航联合基金项目(U1433116)资助;江苏省"333"高层次人才工程;航空科学基金项目(20145752033)资助

摘　　要：不同软件可靠性模型对可靠性数据差异存在适用性的区别.可靠性数据差异可由数据异常、数据结构和数据容量三个方面度量,其中数据异常可通过分层聚类等方法降低.目前国内外尚未出现针对可靠性数据差异,比较模型评测性能和适用程度的工作.鉴于此,本文从上述差异的三个角度入手,分析比较6种重要模型对可靠性数据差异的依赖性.基于大量实验得出一些有价值的启示.Each software reliability model has its flexibility on the differences in software reliability data. The differences can be made up of data exception, which can be reduced by some methods such as hierarchical clustering algorithm, data structure and data size. So far,numbers of researches are made,while papers focus on the comparison of models on differences in reliability data can hardly be found. This paper analyzes 6 important models＇ dependency on 3 kinds of differences mentioned above with extensive experiments, from which several empirical insights are obtained.

关 键 词：软件可靠性 可靠性数据差异 可靠性参数 可靠性评测 分层聚类法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]