arXiv 1907.017792019-07-03cs.SE

使用二元决策图进行组合交互测试中的约束处理

Using binary decision diagrams for constraint handling in combinatorial interaction testing

Tatsuhiro Tsuchiya

摘要：

组合交互测试中测试参数之间的约束关系对测试用例生成的性能有着重要的影响。本文研究了使用二元决策图（BDD）的约束处理的有效性。BDD是一种用于表示和操作布尔函数的数据结构。约束处理程序的核心作用是执行检查，以确定具有未指定参数值的部分测试用例是否满足约束。在生成测试套件的过程中，这种检查会被执行多次;因此，检查的效率会显著影响测试用例生成所需的总时间。在本文中，我们研究了两种不同的方法。第一种方法通过计算布尔函数的逻辑与来执行这种检查，布尔函数代表所有满足约束的完整测试用例和给定的部分测试用例。第二种方法使用一种新的技术来构造一个BDD，表示所有的约束满足部分测试用例。使用此BDD，可以通过简单地从根到接收器遍历BDD来执行检查。我们开发了一个程序，将这两种方法到IPOG，一个著名的测试用例生成算法。使用这个程序，我们经验性地评估这些基于BDD的约束处理方法的性能，共使用62个问题的实例。在评估中，这两种方法进行了比较，三种不同的约束处理方法，即基于布尔可满足性（SAT）解决，最小禁止元组（MFTs），和约束满足问题（CSP）解决。评估结果表明，这两个BDD为基础的方法通常优于其他约束处理技术和BDD为基础的方法，使用新的技术表现出最好的性能。

英文摘要：

Constraints among test parameters often have substantial effects on the performance of test case generation for combinatorial interaction testing. This paper investigates the effectiveness of the use of Binary Decision Diagrams (BDDs) for constraint handling. BDDs are a data structure used to represent and manipulate Boolean functions. The core role of a constraint handler is to perform a check to determine if a partial test case with unspecified parameter values satisfies the constraints. In the course of generating a test suite, this check is executed a number of times; thus the efficiency of the check significantly affects the overall time required for test case generation. In the paper, we study two different approaches. The first approach performs this check by computing the logical AND of Boolean functions that represent all constraint-satisfying full test cases and a given partial test case. The second approach uses a new technique to construct a BDD that represents all constraint-satisfying partial test cases. With this BDD, the check can be performed by simply traversing the BDD from the root to a sink. We developed a program that incorporates both approaches into IPOG, a well-known test case generation algorithm. Using this program, we empirically evaluate the performance of these BDD-based constraint handling approaches using a total of 62 problem instances. In the evaluation, the two approaches are compared with three different constraint handling approaches, namely, those based on Boolean satisfiability (SAT) solving, Minimum Forbidden Tuples (MFTs), and Constraint Satisfiction Problem (CSP) solving. The results of the evaluation show that the two BDD-based approaches usually outperform the other constraint handling techniques and that the BDD-based approach using the new technique exhibits best performance.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

摘要：

英文摘要：

相关论文