Software Reliability Management through Metrics

As systems become increasingly software dependent, their reliability will accordingly depend more so on the reliability of their resident software. Just as techniques and processes were developed and improved to ensure hardware reliability, so must techniques evolve to ensure software reliability. Two questions are addressed by this thesis. First, how do we measure software reliability throughout a projects lifecycle Second, is there a tool which will provide effective insight into the test-now-or-later problem The solution to our first question is a U.S. Army software procurement methodology which is briefly outlined as the overall framework for software procurement in this thesis. A software fault analysis tool is developed and programmed. Some results from this algorithm are provided and their potential resource saving impact explored. Program managers of software-intensive projects would be well advised to use the Armys methodology and this fault analysis tool to potentially save critically short procurement resources.... Software Reliability, Software Metrics, Reliability