Simulation and comparison of reliability growth models

Abstract

The Duane reliability growth model has a number of inherent limitations that make it unsuitable for monitoring reliability improvement progress. These limitations are explored and a model based on variance-stabilizing transformation theory is explained. This model retains the ease of use while also avoiding the disadvantages of the Duane model. It represents a more useful graphical model for portraying reliability improvement at development team meetings. Computer simulations have shown that the new model provides a better fit to the data over the range of Duane slopes normally observed during a reliability growth program. The instantaneous mean time between failures (MTBF) equation for the new model is developed. Computer simulations show that its use results in higher values of instantaneous MTBF than that achieved by the Duane model. The new model also reduces the total test time for achieving a particular specified instantaneous MTBF. Finally, software failure data from an actual project illustrates the calculations and benefits of the new model. DuÀinst = The instantaneous MTBF of the Duane model. SqÀinst = The instantaneous MTBF of the new growth model.