Reliability models for very large software systems in industry

Abstract

BNR, the R&D subsidiary of Northern Telecom and Bell Canada, has one of the largest software systems in the world, with code libraries exceeding 8 million source lines of a high level language. This software is used in the high-end digital switching systems that Northern Telecom markets. Software reliability methods are applied to a major subset of this software to determine if the total number of customer-perceived failures and actual software faults can be predicted before or soon after a new release of such a system. These predictions are based on pre-customer testing ( alpha and beta ) and small field trials. Many of the existing reliability models and methods of parameter estimation currently demonstrated in the literature are compared.<>