The Software Static Analysis Reliability Toolkit
Walter Schilling and Mansoor Alam
The 17th IEEE International Symposium on Software Reliability Engineering (Student Program Papers ISSRE 2006)
Raleigh, North Carolina, USA, November 6-10, 2006
Abstract
Software reliability represents an increasing risk to overall system reliability. As systems have become larger and more complex, mission critical and safety critical systems have increasing functionality controlled exclusively through software. It is imperative that practical reliability modeling techniques be developed to address these issues. One such proposed model combines static analysis with limited testing and code coverage measurement to estimate the reliability of existing software. Static analysis is used to detect faults within the source code which may lead to failure. Code coverage is used to determine which paths within the source code are executed as well as how often they execute. Finally, a Bayesian Belief Network is then used to combine these parameters and estimate the resulting software reliability. In order to facilitate the usage of this model, the Software Static Analysis Reliability Toolkit (SoSART) has been proposed. This tool serves as both a reliability modeling tool and a bug finding meta tool suitable for comparing the results of different static analysis tools. This article discusses the purpose for this tool as well as the capabilities of this toolset.