很少有软件成本模型会考虑测试的实际情况,因此不能够准确地反映测试过程中成本的实际消耗以及当测试情况需要发生改变时对软件成本需求的变化。绝大多数的软件测试存在着不完全排错和学习过程等现象,这些现象通常反映了实际测试的成本使用情况。本文提出了一个反映实际测试情况的测试级别,在此基础上构造了一个新的软件成本模型,并讨论了基于最低软件成本的最优发布策略。实验证明该成本模型更符合软件测试的实际情况,软件开发者可以利用该模型实现对测试成本科学的管理。
Abstract
Since few software cost models take into account the actual cost of software tes ting, they can not reflect actual consumption of cost and the needed cost when the testing circumstance changes. Imperfect debugging and learning process exist in most software testing processes which usually reflect the actual cost in sof tware testing process. In this paper, a testing level function which reflects ac tual testing circumstance is proposed, and a new software cost model is develope d based on the testing level function. The optimal release polices to minimize t he expected total software cost are discussed. Numerical experiments are present ed and the results show that software cost model based on the testing level refl ect the consumption of the testing resources more realistically. By using this m odel, software developers can manage testing cost scientifically.
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1]LYU M R. Handbook of software reliability engineering[M]. New Y ork: McGrawHill and IEEE Computer Society, 1996.
[2]PHAM H. Software reliability and cost models: Perspectives, comparison, a nd practices[J]. European Journal of Operational Research, 2003, 149(3): 475-4 89.
[3]PHAM H, ZHANG X. NHPP software reliability and cost models with testing c overage[J]. European Journal of Operational Research,2003,145(2): 445-454.
[4]BOLAND P J, CHUIV N. Optimal times for software release when repair is im perfect[J]. Statistics & Probability Letters, 2007, 3 (4):1-5.
[5]ABU G, CANGUSSU J W. A quantitative learning Model for software testing p rocess[C]∥38th international conference on system science, Hawaii, 2005.
[6]PHAM H. A software cost model with imperfect debugging, random life cycle and penalty cost[J]. International Journal on Systems Science, 1996, 27(5):45 5-463.
[7]HUANG C Y, LYU M R. Optimal release time for software systems considering cost, testeffort, and test efficiency[J]. IEEE Transaction on Reliability, 2005, 54(4):583-591.
[8]XIE M, YANG Bao. A study of the effect of imperfect debugging on software development cost[J]. IEEE Transaction on Software Engineeriog, 2003, 29(5): 471-473.
[9]EHRLISH W, PRASANNA B, STAMPFEL J. Determiningthe cost of a stoptest de cision[J]. IEEE Software, 1993, 33(3):33-42. 〖ZK)〗
[10]ZHANG Xuemei, TENG Xiaolin, PHAM H. Considering fault remova l efficiency in software reliability assement[J]. IEEE Transaction on system, man, and cyberneticspart a: systems and human, January 2003, 33(1):114-120.
[11]YAMADA S, OHTERA H, NARIHSA H. Software reliability growth model with te sting effort[J]IEEE Transaction on reliability, 1986, 35(1):19-23.
[12]PHAM H, ZHANG Xuemei. A software cost model with warranty and risk costs [J]. IEEE Transaction on computers, 1999, 48(1):71-75.
[13]HUANG C, LIN C. Software reliability analysis by considering fault depen dency and debugging time lag[J]. IEEE Transaction on Reliability 2006, 55(3): 436-450.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}