A New Model for Software Reliability

As software being used in a larger scale, the problems of their reliability become more and more important. In addition to the design problems of the software itself, more attention was paid to the statistical law showed by the software reliability. As researches getting along, lots of patterns are raised. However, some of the patterns pay too much attention to the randomness of the software failure processes. And in some cases, software failure processes show more regularity than randomness through tests. According to the above, we put forward a static model with the idea of the power system. And we get a very good result by this method. In this paper, we are not going to deny the randomness of the processes, but want to prove that if we consider randomness on the foundation of the power of the process, we would get a better result.