王凡Wang, Farn臺灣大學:電子工程學研究所江志政Chiang, Chih-ChengChih-ChengChiang2010-07-142018-07-102010-07-142018-07-102009U0001-2907200920592000http://ntur.lib.ntu.edu.tw//handle/246246/189260在軟體系統上設計的困難度隨著這幾年軟硬體的發展越趨困難。在現今大型軟體的開發與設計一再的顯示出軟體驗證與測試愈來愈有其必要性。軟體測試上，我們需要使用模型化的方法產生測試方法與驗證。通常因為使用者所使用的模型常常語意不清，所以我們需要明確地定義模型用以滿足使用者所期望的運行結果。在本篇論文我們利用機率分配的方法執行我們定義好的時間狀態機。當我們進入狀態機裡的某一個狀態下，我們會使用機率分配方法考慮該在什麼時間點離開這個狀態。當要離開某個狀態時候，我們再度使用機率分配方法將可選的下一個狀態再做一次機率分配圖，在離開本狀態的時後選擇要進入的下一個狀態。整個實驗過程我們著重在什麼時間點之下才能傳送測試資料執行測試。我們的測試資料最後將轉換成通訊協定測試語言的格式執行。The Software testing needs the use of a model to guide such endeavors as test selection and test verification. Usually, these implicit models exist only in the head of a human tester. The probability distribution is our concept to run time automaton. When we enter a location, we simulate the leaving time point at one probability distribution. After deciding to leave the location, we also need to know where the next location to enter. Once again, we choose the next location by the method of probability distribution. Because we focus on the time-domain on time automaton, we care about the test case should be set on the time point. There are two types of fault. One happen in duration of time on a location, and another happen in transition. Finally, we transfer the time domain test case to TTCN3 test case.誌謝 i要 iiBSTRACT iiiONTENTS ivIST OF FIGURES viiIST OF TABLES ixhapter 1 Introduction 1.1 Thesis frame 2hapter 2 Related work 3hapter 3 Background 4.1 Model-based Testing 4.2 Probability distribution models 5.3 TTCN-3 11.4 Software Testing Workbench 15hapter 4 Model construction of the specification 17.1 Timed Automaton with Variables 17.1.1 Clock constraints 18.1.2 Variable constraints 18.1.3 Semantics and Syntax 19.1.4 Example 20.2 Model construction of the specification 21.2.1 Specifications 21.2.2 Transfer the specification to data structure 22hapter 5 Procedure of test generation 24.1 Testing goal and measure 24.2 Test plan generation of timed states with shortest path 26.2.1 Procedure of Test plan generation of timed states with shortest path 27.2.2 Digraph 29.2.3 Timed location 29.2.4 Shortest Path between two location 30.2.5 Distribution a coverage from a location in probability distribution 32.3 Test plan generation with probability distribution 35.3.1 Procedure of test plan generation with probability distribution 36.3.2 Using probability distribution to select a transition 38.4 Conclusion of test cases generation 41.4.1 Analysis for test plan generation of timed states with shortest path 42.4.2 Analysis for Test plan generation with probability distribution 42hapter 6 Fault model in SUT 44.1 The bug on state 44.2 The fault on transitions 45.3 Set bugs in SUT 45.4 Software Under Testing 46hapter 7 Experiment and Result 48.1 Operational processes 48.2 Result 54.2.1 Test case for a Timed-state 54.2.2 Test plan for simulation path with transition fault 56hapter 8 Conclusion 58EFERENCE 60ppendix A 62ppendix B 631700169 bytesapplication/pdfen-US軟體測試狀態機時間測試測試自動產生Software behavior modelsfinite state machinestest automationtest case generationthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/189260/1/ntu-98-R96943112-1.pdf