PCM预测软件性能的研究与设计

The research and design on software performance prediction with PCM

为了减少基于组件的软件性能预测技术复杂度,研究了软件性能预测的方法.基于PCM(Palladio Component Model)性能预测方法,分析了PCM与UML的建模关系与PCM建模过程,指出架构级和组件级参数之间的关联.此两类不确定性参数的设置与取值是PCM建模的关键,同时,性能预测的成功需要预测工具的有利支持.最后提出了一种PCM性能预测工具的设计方法.

To reduce the technique complex about component-based software performance prediction, soft- ware performance prediction methods are researched. Based on the PCM method, the modeling relation and progress between PCM and UML are analyzed, the contact between framework and component pa- rameters is presented,the two kinds of imperfect parameters are the key on PCM modeling. The per- formance prediction tool is needed to support the successful prediction. Finally,one design method on the PCM performance prediction tool is provided.