Estimation of Optimal Testing-termination Time with Change-point Derived from Software Release  
Author Yuka Minamino

 

Co-Author(s) Shinji Inoue; Shigeru Yamada

 

Abstract Software development companies, as a strategy, release software early to prevent lost opportunities in its competitive market. After software is released, users participate in the test team to detect faults in the software. This testing and operational phases are conducted in parallel. For these firms, the change-point, which alters the stochastic behavior of the software failure-occurrence time interval, occurs because the testing environment alters at the release time due to participation of users. However, the influence of the change-point derived from software release is not considered for the optimal software release problem. This study assumes the release time as the the change-point occurrence time in the environment, which releases software early and has the testing and operational phases conducted in parallel. We discuss the influence of change-point to the expected total software cost by constructing the cost function from the change-point model. Also, we estimate the optimal testing-termination time to minimize the expected total software cost. Finally, we show application examples using actual data sets.

 

Keywords Software Reliability Growth Model (SRGM), Change-point, Optimal Testing-termination Time, Software Development Management
   
    Article #:  RQD26-92
 

Proceedings of 26th ISSAT International Conference on Reliability & Quality in Design
Virtual Event

August 5-7, 2021