軟體可靠度是指一個給定的軟體，在一個明確的環境及時間週期內，能正確無誤地執行所賦予的任務與功能的機率。基於程式編碼和模擬與測試的統計分析與最佳化搜尋之需求，我們選用SAS （Statistical Analysis System）電腦程式語言去建構軟體可靠度評估系統。現有SAS/CARE軟體可靠度評估系統包括下述八個模型之電腦化程式： Non-homogeneous Poisson Process（NPP）Model、General Poisson（GP）Model、Binary Markov Process（BMP）Model、Non-Uniform Path-Selection（NUP）Model、Random Paths（RP）Model、Piecewise Linear Hazard Rate（PLHR）Model、Lognormal Proportional Model（LPM）和Non-Parametric Software Reliability Growth （NPSRG）Model。

　　本研究再加入Non-homogeneous Poisson Process Models的整合性（Unified）模型到此系統中，即Goel-Okumoto模型、Gompertz Growth Curve、Logistic Growth Curve、Generalized Goel NHPP模型、Delayed S-Shaped模型、Inflected S-Shaped模型、Modified Duane模型、Two-Error-Type模型、Weibull-Type Testing-Effort Function模型和Log-Logistic Software Reliability Growth模型。增加模型到原有的軟體可靠度評估系統有四個步驟，第一，模型分析，包括模型分類和模型選擇。第二，參數估計式推導，運用統計上的估計方法進行參數估計式推導。第三，程式編碼，以電腦語言SAS為工具，將上一階段所推導的演算法進行程式實作。第四，模擬與測試，以模擬產生的失效資料，進行測試可得未知參數之估計值，並運用一些數值分析方法進行最佳化（Optimization）求解程序。擴增後的SAS軟體可靠度評估系統將提供軟體品保人員在評估與預測軟體之可靠度時，有更多選擇。

　　Software reliability is defined as the probability of failure-free software operation for a specified period of time in a specified environment. Based on the statistics of coding, simulation, and testing, and the requirement of optimal search, we choose SAS (Statistical Analysis System) program to built the software reliability evaluation system. Existing SAS/CARE software reliability evaluation system has implemented Non-homogeneous Poisson Process (NPP) Model, General Poisson (GP) Model, Binary Markov Process (BMP), Non-Uniform Path-Selection (NUP) Model, Random Paths (RP) Model, Piecewise Linear Hazard Rate (PLHR) Model, Lognormal Proportional Model (LPM), and Non-Parametric Software Reliability Growth (NPSRG) Model.

　　This research further implements the unified models of Non-homogeneous Poisson Process Models, including Goel-Okumoto Model, Gompertz Growth Curve, Logistic Growth Curve, Generalized Goel NHPP Model, Delayed S-Shaped Model, Inflected S-Shaped Model, Modified Duane Model, Two-Error-Type Model, Weibull-Type Testing-Effort Function Model, and Log-Logistic Software Reliability Growth Model. There are four main steps to add models to original software reliability evaluation system. First, the model is reviewed to clarify assumptions, including the classification and selection. Then, estimators of model parameters are obtained. In the third stage, SAS/IML programs are coded. The fourth stage carries out tests using simulation data and applies some numerical analysis to the optimization. The expanding SAS software reliability evaluation system can help software engineers have more decisions when evaluating and forecasting the reliability.

[1]Abdel-Ghaly, A. A., Chan, P. Y., and Littlewood, B., Sept. 1986, “Evaluation of competing software reliability predictions,” IEEE Transactions on Software Engineering, Vol. SE-12, pp. 950-967.

[2]Angus, J. E., Schafer, R.E., and Sukert, A, Jan. 1980, “Software reliability model validation,” Proc. Annu. Reliability and Maintainability Symp. , San Francisco, CA., pp. 191-199.

[3]Brooks, W. and Motley R., 1980, Analysis of Discrete Software Reliability Models, Tech. Report RADC-TR-80-84, Rome Air Development Ctr, Rome, N.Y.

[4]Chen, S. and Mills, S., Mar. 1996, “A binary Markov process model for random testing,” IEEE Transactions on Software Engineering, Vol. 22, No. 3, pp. 218-223.

[5]Crow L., 1977, “Confidence Interval Procedures for Reliability Growth Analysis,” Tech. Report 197, US Army Materiel Systems Analysis Activity, Aberdeen, Md.

[6]Crow, L. H., 1974, “Reliability Analysis for Complex, Repairable System in Reliability and Biometry, ” F. Proschan and R. J. Serfling. Philadelphia. PA: SIAM, pp. 379-410.

[7]Downs, T. and Garrone, P., Aug 1991, “Some new models of software testing with performance comparisons,” IEEE Transactions on Reliability, Vol. 40, No. 3, pp. 322-328.

[8]Duane, J. T., 1964, “Learing curve approach to
Reliability Monitoring,” IEEE Transactions on Aerosp., Vol. 2, pp. 563-566.

[9]Gaudoin, O., Lavergne, C., and Soler, J. -L., Dec 1994, “A generalized geometric de-eutrophication software-reliability model,” IEEE Transactions on Reliability, Vol. 43, No. 4, pp. 536-541.

[10]Goel A. and Okumoto K., Aug. 1979, “Time-Dependent Error-Detection Rate Model for Software Reliability and Other Performance Measures,” IEEE Transactions on Reliability, pp. 206-211.

[11]Goel A. L., Dec. 1985, “Software Reliability Models: Assumptions, Limitations, and Applicability,” IEEE Transactions on Software Engineering, Vol. SE11, NO. 12.

[12]Gokhale S. S. and Trivedi K. S., 1998, “Log-logistic Software Reliability Growth Model,” Proceedings of the 3rd IEEE Internation High-Assurance Systems Engineering Symposium, pp. 34-41, Nov. 13-14, Washington, DC.

[13]Huang C. Y., Lyu M. R., and Kuo S. Y., 2003, “A Unified Scheme of Some Nonhomogenous Poisson Process Models for Software Reliability Estimation, ” IEEE Transactions on Software Engineering, Vol. 29, pp. 261-269.

[14]Jelinski Z. and Moranda P., 1972, ”Software Reliability Research,” in Statistical Computer Performance Evaluation, W. Freiberger, Academic Press, New York, pp. 465-484.

[15]Keiller P. et al., 1983, “Comparison of Software Reliability Predictions,” Proc. IEEE Int’l Symp. Fault-Tolerant Computing, IEEE CS Press, Los Alamitos, Calif., pp. 128-134.

[16]Littlewood B., Oct. 1981, “Stochastic Reliability Growth: A Model for Fault Removal in Computer Programs and Hardware Designs,” IEEE Transactions Reliability, , pp. 313-320.

[17]Littlewood B. and Verrall J., “A Bayesian Reliability Growth Model for Computer Software,” F. Royal Statistics Soc. C, Vol. 22, pp. 332-346.

[18]McCabe, T. J., 1976, “A Complexity Measure,” IEEE Trans. Software Engineering, Vol. SE-2, No. 4.

[19]Miller D., Jan. 1986, “Exponential Order Statistic Models of Software Reliability Growth,” IEEE Transactions Software Eng., pp. 12-24.

[20]Moranda P., Dec. 1979, “Event-Altered Rate Models for General Reliability Analysis,” IEEE Transactions Reliability, pp. 376-381.

[21]Musa J. and Okumoto K., 1984, “A Logarithmic Poisson Execution Time Model for Software Reliability Measurement,” Proc. Int’l Conf. Software Eng., IEEE CS Press, Los Alamitos, Calif., pp. 230-238.

[22]Ohba M., July 1984, “Software Reliability Analysis Models,” IBM J. Res. Develop., Vol.28, NO. 4, pp. 428-443.

[23]Pham, H., 2000, Software Reliability, Springer, New York.

[24]Pham, H. and L.Nordmann, 1997, A generalized NHPP software reliability model, in Third International Conference on Reliability and Quality in Design, Anaheim, ISSAT Press.

[25]Ross S. M., 1996, “STOCHASTIC PROCESSES, ”Wiley.

[26]Schneidewind N., June 1975, “Analysis of Error Processes in Computer Software,” SigPlan Notice, pp. 337-346.

[27]Schafer P. et al., 1979, “Validation of Software Reliability Models,” Tech. Report RADC-TR-79-147, Rome Air Development Ctr., Rome, N.Y..

[28]Sofer, A. and Miller, D.R., Aug 1991, “A nonparametric software-reliability growth model,” IEEE Transactions on
Reliability, Vol. 40, No. 3, pp. 329-337.

[29]Yamada S. and Osaki S., 1984, “Non-homogeneous Error Detection Rate Models for Reliability Growth,” Stochastic Models in Reliability Theory, eds, S. Osaki and Hatoyama, Berlin Springer-Verlag, pp. 120-143.

[30]Yamada S., Hishitani J, and Osaki S., April 1986, “Software Reliability Growth Models with Testing Effort, ” IEEE Transactions on Reliability, Vol. R-35, No. 1, pp. 19-23.

[31]Yamada S., Ohba, M., and Osaki, S., Dec. 1983, “S-Shaped Reliability Growth Modeling for Software Error Detection,” IEEE Transactions Reliability, pp. 475-478.

[32]Xie, M., 1991, Software Reliability Modeling, World Scientific Publishing Company.

[33]黃佳雄，2002年6月，「建構包含四種模型之軟體可靠度評估系統」，碩士論文，國立成功大學製造工程研究所。