Modelling software operational reliability via input domain-based reliability growth model

Operational reliability of programs depends upon many factors imposed by design, testing, and operation of the programs. In the testing stage, the main imperfect factors which can affect operational reliability are faults, testing time, fault correction, testing profile and operation profile. The proposed modelling process takes these factors into account. It consists of two testing stages and one operational stage. In the testing stages, partition testing and an input domain-based reliability growth model are used to evaluate the reliability growth. Based on the profile coverage between testing and operational profiles, a method is developed to partition the input domain, so that the testing profile can follow the given operational profile of a program. In the operational stage, the proposed reliability model uses the remaining faults from testing stages as the source of unreliability. A case study is conducted using the proposed models.