摘要： 针对iptables核心包分类算法的低效问题，提出一种符合Linux内核限制条件并充分利用已有内核机制的高效包分类算法。该算法具备动态更新、多维匹配、实施速度快等主流包分类算法的特点，适合实际应用。实验结果表明在规则库较大的情况下，算法性能有很大提高。
关键词:
范围定位,
包分类,
Linux内核
Abstract: The iptables’ core packet classification algorithm has lower capability. Focusing on this issue, this paper presents an efficient packet classification algorithm according to the limits of the Linux kernel, as well as sufficient usage of the mechanisms in the kernel. It provides the characteristics, such as multi-dimensional packet classification, dynamic operations and quick matching, which are required by the mainstream. Experimental results show that the algorithm greatly improves the performance, especially in the condition of big rule sets.
Key words:
range location,
packet classification,
Linux kernel
中图分类号:
雷渊明;张大方;谢 鲲;何施茗;杨 虎. 一种基于Linux内核的高效包分类算法[J]. 计算机工程, 2009, 35(20): 261-263.
LEI Yuan-ming; ZHANG Da-fang; XIE Kun; HE Shi-ming; YANG Hu. Efficient Packet Classification Algorithm Based on Linux Kernel[J]. Computer Engineering, 2009, 35(20): 261-263.