New approach to determine the critical number of failure in software systems

Software-Engineering is very important today. In industry (specifically by software critical system) it is important to produce high reliable software, i.e. software with low proportion of faults. To produce such reliable software, a long handling process is required, and because this process consumes a large amount of time and resources to achieve the desired reliability goals it is useful to use Software Reliability Stochastic Models to predict the required software testing time. In this paper a new approach to reflecting the residual number of critical failures in software-systems is introduced. There are currently very few processes enabling us to predict the reliability of the critical failures or the critical failure rate for critical systems. Furthermore, we will focus on distinguishing the critical failures in the software. We will thus distinguish both critical as well as non-critical failures in the Software. Therefore it is important to divide the process into two classes, detection- and correction class. To develop an approach it is necessary to determine corresponding distribution functions and model assumptions.