利用决策树实现包过滤的关键技术

Key Technology of Packet Filter Using Decision Trees

把决策树分类器引入包过滤技术当中，改变了传统的顺序检索包过滤的方法。以 增加预处理时间为代价，提高了数据包的过滤速度。该文对决策树分类器的建树算法、搜索 算法以及可递增性等关键技术进行了研究，并通过仿真实例和性能分析表明了该方法的有效 性。

Introducing the decision tree classifiers into packet filtering tec hnology, it changes the conventional method searching the packet orderly. The ne w technique improves the filtering speed at the cost of preprocess time. Some ke y technologies such as the building and searching algorithms and the ability of increasable by degrees are researched. The instances and performance analysis pr ove the validity of this method.