Software Reliability Prediction Method Based on Composite Poisson Process

Abstract

The invention discloses a compound poisson process-based software reliability prediction method, and belongs to the field of software reliability prediction. The method is characterized in that the method is used for performing strong reliability prediction based on cost and profit loss during the operation and the maintenance of a software according to a set loss level. The method comprises the following steps of performing prediction within the range of a set number of prediction failure data sets based on a plurality of historical data sets consisting of data sets at an equal time interval, wherein each data set consists of the starting time and the ending time of the operation of the software, accumulated failure loss and accumulated loss magnitude; gradually predicting accumulated loss magnitude from a software publication date to a current prediction date according to previous accumulated loss magnitude by taking the ending time of the last historical data set as a prediction starting value and considering the accumulated loss magnitude as a random process consistent with a random process from the software publication date to the current prediction date; and performing comparison with historical data in one-to-one correspondence with predicted data in each prediction time within a set error range. The reliability problem of the software under different operating conditions is solved.

Description

technical field [0001] The present invention belongs to an important attribute of software credibility, that is, a measurement model of software reliability, and specifically relates to a G-O software reliability model and a compound Poisson process to integrate the severity of software failure into software reliability measurement and improve the comprehensiveness of measurement , a unified measure of software credibility. Background technique [0002] With the rapid development of the Internet and information technology, the information infrastructure with communication, storage and computing as the core has penetrated into all aspects of politics, economy, military, culture and social life. At the same time, due to the huge loss caused by software failure, various countries, organizations and researchers began to pay attention to how to ensure the correct and effective operation of software, that is, the credibility of software. [0003] Software reliability is generally...

AI Technical Summary

Problems solved by technology

Different attributes have different measurement scales. For example, reliability is the failure time, security is the number of vulnerabilities, and applicability is the percentage of satisfaction. The scale is not uniform, which makes it difficult to measure the current software credibility. It is achieved by comparing subjective expert scoring methods. Therefore, a unified measurement scale is urgently needed to unify the measurement of different credible attributes.

Images