The model for software quality measurement, using the "genetic" feature of the software

The author of this paper believe, that software system by itself is a complex object, which could be described via the Linear Mixed Model ANOVA. It includes different parameters, which could have fixed (non-random) or stochastic nature. There is produced an idea about software quality prediction by defect number using "genetic" feature of the software (Kokina 2003). It is believed that the term of "genetic" feature is an initial software feature of the software affecting the software from the moment it was initially developed. There are presented a mathematical model for "genetic" quality of software evaluation by Best Linear Prediction (BLP) method and some results of the experiment using simulation of this mathematical model in MATLAB here (Kokina and Petersons 2004). Finally it is presented an idea of real complex software system simulation and a reference to the produced mathematical model for evaluation of the real software "genetic" feature in this paper. Currently the author is working at this idea and software model simulation. This paper is a report of the current stage of the exploration work in the field of the software quality evaluation using the term of the "genetic" feature and simulation of produced model and the real application system. The previous report was presented in (Kokina 2004).