The Goel-Okumoto (1979) model of reliability growth is applied to the failure data for a large software development project. This is done by fitting the model parameters to the error counts and test times, which are periodically tabulated. In the early phases of system testing, the problem of inaccurate estimation due to insufficient reliability growth occurs. A simple test is devised to measure the stability of the data with respect to this problem. In practice, such a test on the data can be used to determine when system testing has sufficiently progressed to allow one to expect reasonable estimates from reliability growth models.>

Multiprocessor systems find wide application in telecommunications. Simulation techniques are often incorporated in the performance analysis of these systems. Simulation models enable an operating system to be represented in detail, e.g., multitasking, message handling, and concurrent processing. This level of detail is useful in evaluating software design. However, the complexity of such models can lead to costly simulations. Aggregation techniques can often be used to reduce model complexity. Aggregation typically involves the representation of only part of the system in detail, i.e., the primary subsystem. The remainder of the system is represented by a simplified composite. Parametric analysis is performed on the primary subsystem. The role of the composite is to introduce interactive effects to the primary subsystem. Aggregation techniques are accurate for certain classes of analytic models. The purpose of this study is to apply aggregation to the simulation model of a multiprocessor switching system. A model for real-time performance evaluation is described. An aggregation technique is derived from basic queueing models. The accuracy of the aggregated model is assessed through the simulation of realistic test cases. The results demonstrate that good accuracy can be obtained with aggregation. Substantial simulation cost savings are also demonstrated.