随着传感器技术的发展，运行在各种移动设备上的上下文感知应用也迅速增加。这些应用软件不断地捕获移动设备周围的环境信息作为上下文，然后自动地调整设备的状态来适应上下文的变化。然而它们适应的过程中可能会由于触发了错误的适应规则而出错，因此，对这些上下文感知应用进行测试是保证其发展的至关重要的先决条件。由于上下文状态的多样性，在设计测试用例的时候可能会存在组合爆炸问题。论文了提出了一种基于正交试验设计的测试用例选择方法，并开发了一个普通的上下文感知应用来证明该方法对测试用例选择的有效性。此外，该文还对实验数据进行分析来展现该方法所特有的错误定位能力。实验结果表明，该方法不仅能够有效地选择出高效的测试用例，而且还能够利用这些测试用例来定位一个上下文感知应用的程序错误。%Context-aware applications(CAAs) running on mobile devices are now developing quite rapidly due to the ad-vanced built-in sensor technologies .These applications constantly capture environmental information as contexts and then a-dapt to the contextual changes .However ,their adaptations may contain errors when erroneous adaptation rules are trig-gered .Hence ,testing these applications becomes a vital prerequisite for their development .Due to the diversity of contexts , there may be a combinatorial explosion problem during test case generating .This paper proposes a new optimization method for selecting efficient test cases from the perspective of orthogonal experiment design (OED) .A common context-aware ap-plication is developed to prove the effectiveness of this method .Additionally ,an analysis of the fault localization capabilities is presented .Experiment results show that our method can not only significantly select out efficient test cases but also find out the factors that cause the faults of a CAA .
展开▼
