|
For Full-Text PDF, please login, if you are a member of IEICE,
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
|
Software Profit Model under Imperfect Debugging and Optimal Software Release Policy
Chong-Hyung LEE Kyung-Hyun NAM Dong-Ho PARK
Publication
IEICE TRANSACTIONS on Information and Systems
Vol.E85-D
No.5
pp.833-838 Publication Date: 2002/05/01 Online ISSN:
DOI: Print ISSN: 0916-8532 Type of Manuscript: PAPER Category: Software Engineering Keyword: imperfect debugging, optimal software release policy, profit model, random life cycle, penalty cost,
Full Text: PDF>>
Summary:
This paper considers a software reliability model which allows for two types of imperfect debuggings at each failure of the software system. For one type of imperfect debugging, a fault that causes the failure is imperfectly debugged without altering the fault contents of the software system. For the other type of imperfect debugging, the fault is not only imperfectly debugged, but also a new fault is generated and introduced into the system. The probability of perfect debugging is assumed to be an increasing function of the number of debuggings performed prior to the current failure of the system. Based on the software reliability model presented, we consider three profit models to determine the optimal software release times which maximize the expected software profit. These models consider: (1) constant life cycle, (2) random life cycle, (3) random life cycle and penalty cost which is imposed when the software is delivered late. The optimal release times are shown to be finite and unique. Numerical examples are provided for illustrative purposes.
|
|