论文部分内容阅读
可测试性度量通常是不明确的,且难以选择,在分析可测试性的许多方法中,都是试图计算为了激活网络的一个故障并将其影响传播到一个可观察的输出所需要的结点数目,然后用可控制性和可观察性数目去度量测试生成有多困难.然而,这个度量没有考虑根据搜索一组有效的赋值的反跟踪运算的结点数目(因而包含了计算的作用),因此它并不直接对应测试生成的成本.罗马空军研究中心提出一个分级的、面向成本的可测试性模型.这个模型的核心由一组真实的可测试性度量组成,即:
Testability metrics are often ambiguous and difficult to select. In many approaches to analyzing testability, there are attempts to compute the number of nodes needed to activate a fault in the network and propagate its influence to an observable output However, this measure does not take into account the number of nodes (and hence the computational effects) that are involved in the anti-tracking operation based on searching a set of valid assignments, So it does not directly correspond to the cost of test generation.Ruman Air Force Research Center proposed a hierarchical, cost-oriented testability model.The core of this model consists of a set of true testability measures, namely: