Calibrating Software Reliability Models When the Test Environment Does not Match the User Environment

In our computer and Internet age, software failures have a significant impact on the qualtiy of service and down time of various systems. For many applications, estimating the software reliability from a user's perspective plays a contributing role in the decision of when to release the software. Traditionally, software reliability growth models are applied to the system test data with the hope of estimating the software failure rate in the field. Given the aggressive nature by which the software is exercised during system test, as well as unavoidable differences between the test environment and the field environment, the resulting estimate of the failure rate will not typically reflect the user-perceived failure rate in the field. The goal of this work is to quantify the mismatch between the system test environment and the field environment.