Highly Reliable Software Reliability Assessment Based on Statistics of Extremes and Bootstrapping Method

Abstract:

Reliability assessment of Highly Reliable Software is significant in the software reliability engineering because of the small-size failure data. A novel model based on bootstrapping method and statistics of extremes for highly reliable software reliability assessment was presented. Correlation coefficient method was applied in order to determine the extreme distribution pattern to which the failure data belongs. The bootstrapping method based on residual error was used to estimate the parent distribution parameters. Software reliability and mean-time-to-failure (MTTF) at the end of reliability test were assessed. Experimental results show the model has a higher accuracy in the small-size sample situation. The validity of the proposed method is examined.

• Periodical: Applied Mechanics and Materials (Volumes 670-671)
• Pages: 1477-1481
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.670-671.1477
• Online since: October 2014
• Authors: Hao Nan Tong*, Qiu Ying Li
• Keywords: Bootstrapping Method, Failure Data, Highly Reliable Software, Reliability Assessment, Small-Size Sample, Statistics of the Extremes

References

[1] Kaufman L M, Bechta Dugan J, Johnson B W. Using statistics of the extremes for software reliability analysis[J] Reliability, IEEE Transactions on , 1999, 48/3: 292-299.

DOI: 10.1109/24.799901

[2] Lu Minyan. Software reliability engineering[M]. Beijing: National Defense Industry Press, 2011: 225-244.

[3] Efron B. Bootstrap methods: another look at the jackknife [J]. The annals of Statistics, 1979: 1-26.

DOI: 10.1214/aos/1176344552

[4] Kaneishi T, Dohi T. Parametric bootstrapping for assessing software reliability measures [C]/ 2011 IEEE 17th Pacific Rim International Symposium on, 2011, 1/9: 12-14.

DOI: 10.1109/prdc.2011.10

[5] Inoue S, Yamada S. A Bootstrap Method for Software Reliability Assessment Based on a Discretized NHPP Model[C]/ Recent Advances in Software Dependability (WDTS-RASD), 2012, 23/27: 18-1.

DOI: 10.1109/wdts-rasd.2012.14

[6] Yanqing Wang, Zhongyu Wang. Gray bootstrap method for estimating frequency-varying random vibration signals with small samples[J] Chinese Journal of Aeronautics, 2014, 27/2: 383-389.

DOI: 10.1016/j.cja.2013.07.023

[7] Shi Daoji. Practical extreme statistical method[M]. Tianjin: Tianjin Science and Technology Press, (2006).

[8] Liu Jing. Statistics of extremes and application in financial risk management[D]. Tianjin: Tianjin University.

[9] Kaufman L M, Bechta Dugan J, Johnson B W. Using statistics of the extremes for software reliability analysis of safety critical systems[C]/. The Ninth International Symposium on, 1998: 355-363.

DOI: 10.1109/issre.1998.730901

[10] Chen Yun, Zhang Zhihua. Using statistics of extremes for highly reliable software reliability analysis[J]. Journal of naval engineering university, 2005, 02: 95-99.

[11] Kaufman L M, Smith D T, Bechta Dugan J, Johnson B W. Software reliability analysis using statistics of the extremes[C]/ Reliability and Maintainability Symposium. 1997, 175/180: 13-16.

DOI: 10.1109/rams.1997.571701

[12] Chen Xiru. An introduction to mathematical statistics[M]. Beijing: Science Press, 2007: 550-553.