Study on A Priori Bayesian Demonstration Testing Method for Unknown Software Reliability Index Requirement
Zhenyu Ma, Fusheng Liu, Wei Zhang, Jianping Wang, Kun Han
Abstract
Zhenyu Ma, Fusheng Liu, Wei Zhang, Jianping Wang, Kun Han
Abstract
Software reliability demonstration testing usually sets the software reliability index (failure probability) in advance, and then carries out the checking and acceptance work for the software. However, in the course of some reliability demonstration testing, the software actual reliability level still needs to be determined when no specific software reliability index is proposed. So, in order to solve this problem, a approach of prior Bayesian software reliability demonstration testing based on Fibonacci iteration algorithm is presented according to the Bayesian theory and Fibonacci search principle. Through cases analysis, Fibonacci iterative algorithm can efficiently determine the software actual failure probability. At the same time, this means can significantly reduce the number of test cases under the condition of guaranteeing the same confidence.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Software reliability demonstration testing usually sets the software reliability index (failure probability) in advance, and then carries out the checking and acceptance work for the software. However, in the course of some reliability demonstration testing, the software actual reliability level still needs to be determined when no specific software reliability index is proposed. So, in order to solve this problem, a approach of prior Bayesian software reliability demonstration testing based on Fibonacci iteration algorithm is presented according to the Bayesian theory and Fibonacci search principle. Through cases analysis, Fibonacci iterative algorithm can efficiently determine the software actual failure probability. At the same time, this means can significantly reduce the number of test cases under the condition of guaranteeing the same confidence.
Key concepts: Software quality, Computer science, Reliability engineering, Software reliability testing, Reliability (semiconductor), Bayesian probability, Software, Data mining