软件可靠性测试的理论分析

目前软件可靠性测试没有真正开展起来,即使重要的军用软件也没有进行软件可靠性测试。软件可靠性模型的不一致性以及软件可靠性测试时间长、费用高、资源消耗大是造成这种局面的两个主要问题。而这两个问题很大程度是由传统基于随机系统假设的软件可靠性理论导致的。因此,在该理论下对可靠性测试方法及可靠性模型的修补与改善工作很难从根本上解决问题。只有从软件可靠性理论本身入手,提出新的研究思路,才有可能摆脱目前软件可靠性测试的困境。首先介绍了软件可靠性测试的现状,然后重点对现有可靠性理论进行了分析和总结,在此基础上,提出了软件可靠性测试研究的一些新思路。

Analysis on Software Reliability Testing Theory

Software reliability testing has not been carried out effectively.Even critical military software doesn't carry out software reliability testing.The inconsistency of the software reliability model and the high resources consumption in software reliability testing are two main reasons for the situation,which is mainly caused by the software reliability theory based on the traditional hypothesis of random system theory.Revision and improvement on reliability testing method or reliability model under such theory are hard to solve the problem thoroughly.Only starting with the reliability theory itself,bringing forward new ways,will it be possible to break through the difficulties of software reliability testing.At first,the status of software reliability testing is introduced and then the existing software reliability theories are analyzed.On the basis,new research ideas of software reliability testing are brought forward.