Application on Software Reliability Modeling by Wavelet De-noising
|School||Hebei University of Engineering|
|Course||Applied Computer Technology|
|Keywords||Software reliability model Wavelet Denoising Application Consistency|
With the increasingly wide range of software applications and importance of increasing awareness of software quality requirements are also increasing . Reliability is important as a measure of software quality characteristics, and its quantitative assessment and prediction has become the focus of attention and study . Reliability software reliability evaluation model as the core and key that can be used at different stages of the software life cycle , quantitative behavior of software reliability estimation and forecasting , resource allocation for software , software marketing decisions of great significance. While software reliability models have been developed from the research stage to the engineering stage, but the face of the software development process itself and its increasingly complex situations , it still presents its own deficiencies . One of the most major application of the model is the inconsistency . Because software failure data sequence in the data collection process by a variety of human and non-human factors , making the observed data sequence is always mixed with some noise. The presence of noise seriously interfere with the true colors of the signal is not conducive to failure data for further analysis and processing sequence . Therefore , to eliminate or reduce the noise , in order to maximize the extraction of the useful signal is necessary . It directly affects the follow-up results. Traditional software reliability models in the modeling process often does not consider the characteristics of software failure data itself , ignoring the failure data the presence of noise . In this paper, to solve this problem , the use of wavelet thresholding algorithm software failure data sequence denoising pretreatment ; and the analysis of the advantages and disadvantages of conventional thresholding algorithm is proposed based on an improved thresholding algorithm, new algorithm to some extent make up the traditional thresholding algorithm deficiencies and improve the denoising results. In this paper, three sets of classic software failure data sequence , for example, on the use of a new threshold denoising method for modeling the failure data sequence , and for the excellent degree of fit , predictive ability of the statistical results of step with the model established before denoising compared through calculation, simulation tests confirmed before modeling the data de-noising preprocessing is feasible, effective, and some improvement in the application of software reliability model inconsistencies .