Authors: Marcus A. Agustin, Ma. Zenia N. Agustin
This paper considers a series system of m software where the failure of each soft- ware follows a generalized Jelinski-Moranda model. Whenever the system breaks down, the software that caused the failure under- goes debugging. The imperfect debugging scheme assumes a speci ed probability func- tion p(t) for completely removing the bug responsible for system failure. At the con- clusion of a xed observation and debugging phase, we assess system reliability by assign- ing tasks with random completion times. We will show that we can derive some existing debugging models from the proposed model.
Keywords: Competing risks, imper- fect debugging, software reliability.