Premium
A unified approach of testing coverage‐based software reliability growth modelling with fault detection probability, imperfect debugging, and change point
Author(s) -
Chatterjee Subhashis,
Shukla Ankur
Publication year - 2019
Publication title -
journal of software: evolution and process
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 29
eISSN - 2047-7481
pISSN - 2047-7473
DOI - 10.1002/smr.2150
Subject(s) - debugging , reliability engineering , computer science , software quality , ranking (information retrieval) , reliability (semiconductor) , fault detection and isolation , software , fault (geology) , process (computing) , point (geometry) , data mining , software development , engineering , machine learning , artificial intelligence , programming language , mathematics , power (physics) , physics , geometry , quantum mechanics , seismology , actuator , geology
This paper presents a unified approach to model the reliability growth of software with imperfect debugging and coverage factor. Existing testing coverage‐based software reliability growth models considered that faults present at a particular fault location are detected with certainty during the testing process. Practically, it is very difficult to detect all software faults. To overcome this limitation, a revised software reliability growth model has been developed with the assumption that detection of the faults at a particular fault location is not definite. Furthermore, a new method to model the imperfect debugging phenomenon has been incorporated in the proposed study. A revised model ranking method has been developed to improve the accuracy of model ranking, which is mainly extension of existing normalized criteria distance method. Change point analysis has been done with the effect of different environmental factors on the models' parameters. Numerical examples are given to demonstrate the effectiveness of the proposed model.