Software Quality Assurance Based on Order Statistics

Statistical Process Control is an effective technique to optimize the quality and productivity of large scale software firms. Control charts are extensively used to monitor the process to note the variations in the software process that are results of unpredictable causes that behave in an unintended way results in fixing the bug in a flash by the team lead. For an effective monitoring of failure process the time between every r th failure (r is a natural number >=2) instead of inter-failure times is considered for developing a variable control chart called Time Control Charts. This paper projects a controlling framework based on order statistics of the cumulative quantity between observations of time domain failure data using mean value function of Logarithmic Poisson Execution Time Model (LPETM), which is a Non Homogenous Poisson Process (NHPP). The two unknown parameters of the Logarithmic model are arrived at, using The Maximum Likelihood Estimation (MLE).