Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17–18, 2017, Proceedings

Research Article

Prediction Model Based Failure Time Data for Software Reliability

  • @INPROCEEDINGS{10.1007/978-3-319-73317-3_32,
        author={Peng Lin and Xu Tian and Xiaojuan Wang and Xu Cao and Jiejing Cao and Jianli Li and Yan Gong},
        title={Prediction Model Based Failure Time Data for Software Reliability},
        proceedings={Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17--18, 2017, Proceedings},
        proceedings_a={ADHIP},
        year={2018},
        month={2},
        keywords={Defect prediction Poisson process Data fitting Software reliability},
        doi={10.1007/978-3-319-73317-3_32}
    }
    
  • Peng Lin
    Xu Tian
    Xiaojuan Wang
    Xu Cao
    Jiejing Cao
    Jianli Li
    Yan Gong
    Year: 2018
    Prediction Model Based Failure Time Data for Software Reliability
    ADHIP
    Springer
    DOI: 10.1007/978-3-319-73317-3_32
Peng Lin,*, Xu Tian1, Xiaojuan Wang1, Xu Cao1, Jiejing Cao1, Jianli Li1, Yan Gong1
  • 1: China Electronic Equipment of System Engineering Institute
*Contact email: paul-lim@163.com

Abstract

Since all the defects cannot be detected within a finite software testing process (STP), the failure data should be wisely used to estimate the potential defects for software reliability. Therefore, a standard graphical methodology (GM) model is proposed for software reliability, in which failure data of time domain is utilized to predict the potential defects. First, non homogeneous and compound Poisson process is involved to model the failure time during STP. Then, GM model is utilized to predict the potential defects. Further, the software reliability is estimated based on GM model. Finally, compared with the traditional models, GM model can reach an improvement of 30% relative gain on average.