面向深度分组检测的高速数据分组解析结构

摘要

A deep packet inspection oriented high speed packet parsing architecture called BiPPCS (bidirectional packet parsing architecture for content security) was proposed. Firstly, the content extraction tree was used to describe the coupling of the protocol relationship to improve flexibility of the packet parsing. Secondly, hardware bi-directional paral-lel pipeline was used to enhance the processing rate of the packet parsing. Thirdly, a node mapping algorithm was used to balance the number of nodes on all pipeline stages to optimize the storage space. Analysis and simulation show that BiPPCS gets balance among the rate processing, resource consumption and other aspects.%　　提出了一种面向深度分组检测的高速数据分组解析结构BiPPCS（bidirectional packet parsing architecture for content security）。结构采用内容萃取树描述协议的耦合关系从而提高了数据分组解析的灵活性；利用硬件双向并行流水线提升了数据分组解析的处理速率；通过使用节点映射算法来均衡各级流水线上的节点数目优化存储空间；分析和仿真显示BiPPCS在处理速率、空间利用率等方面能取得较好的均衡。