Abstract: This paper discussed a multi-fractal based method to analyze the fluctuation of the parameters of system resource, and proposed a new methodology which combined qualitative analysis with ctuantitative analysis to predict resource consumption and the trend of software aging. Firstly, this study used fractal theory to discuss the fractal structure of the parameters of system resources that influenced the performance of software system. And the results show that the variations of the parameters are not a stochastic process, but have characteristic properties of fractal. In addition, the characteristics of the spectra can be used to analyze the changes of system parameters during the running time qualitatively. Secondly, this paper put forward a new methodology for calculating multidimensional exponent, which is applied to data of system resource usage. Thirdly, the Auto-Regressive Moving-Average (ARMA) model was adopted in order to carry out the analysis of the multidimensional exponent and build the corresponding forecast model. Finally, the experiment was taken to calculate the multidimensional exponent of parameter series related with several memory resources using the parameters data collected from a realistic software system. The results of the experiment indicate effective ability to predict software aging.

Key words: Software aging,Multi-fractal,Prediction