Testing‐Coverage‐Based Software Reliability Growth Model with Imperfect Debugging and Fault Removal Efficiency

ABSTRACT

Reliable software is essential because failures can disrupt critical services, cause financial loss, and reduce user trust. Software reliability depends not only on how faults are detected, but also on how effectively they are removed. Many existing software reliability growth models (SRGMs) treat these processes separately. This work develops a non‐homogeneous Poisson process (NHPP)‐based model that considers testing coverage, imperfect debugging, and fault removal efficiency (FRE). A flexible testing coverage function is used to capture changing fault‐detection behavior during the testing process. Imperfect debugging represents the introduction of new faults during correction. FRE reflects the proportion of detected faults that are successfully removed. The proposed model is validated using two software fault datasets. Its parameters are estimated through least‐squares estimation, and its performance is examined using mean squared error (MSE), mean absolute error (MAE), and the coefficient of determination (). The model is also compared with existing SRGMs. A one‐at‐a‐time sensitivity analysis is performed to study the influence of the testing coverage and debugging parameters. The proposed framework provides a practical way to describe software reliability growth under realistic testing and debugging conditions.