Modeling software-reliability with multiple failure-types and imperfect debugging

Abstract

This paper presents a software reliability model that is based on a nonhomogeneous Poisson process. The major contribution of this model is combining multiple failure types with imperfect debugging. In addition, the paper discusses cost models that can be used to determine the optimal time to be spent debugging. The software reliability model allows for three different types of errors: critical, major, and minor errors. Critical errors are the most difficult to detect and the most expensive to remove. Major errors are moderately difficult to detect and fairly expensive to remove. Minor errors are easy to detect and inexpensive to remove. The model also allows for the introduction of any of these types of errors during the removal of an error. Using the software reliability model developed, we determine the optimal debugging time necessary to minimize costs subject to reliability constraints. A numerical example is provided to illustrate the above techniques. >