侯慧玲



摘 要: 传统位并行多维数据包分类系统的数据信息占据空间过大、分类效果较差。针对上述问题,基于云计算设计一种新的位并行多维数据包分类系统,主要分为硬件设计和软件设计两部分。系统硬件包括三个模块,分别为规则管理模块、存储管理模块和中断处理模块;软件由内部分类程序、数据分类程序、数据包包头合成程序和内核数据处理程序4个主要程序组成。为检测系统的工作效果,与传统系统进行实验研究,结果表明,基于云计算的位并行多维数据包分类系统数据信息占据空间小,分类效果优于传统系统设计。
关键词: 数据包分类; 位并行算法; 系统设计; 云计算; 位并行多维数据包; 内核数据处理
中图分类号: TN99?34; TP394 文献标识码: A 文章编号: 1004?373X(2020)17?0175?05
Abstract: The data information designed by the traditional bit?paralleled multi?dimension packet classification system occupies too large space and the classification effect is poor. In view of this, a new bit?paralleled multi?dimension packet classification system based on cloud computing is designed. The system consists of hardware and software. The system hardware includes three modules, which are rule management module, storage management module and interrupt processing module. The system software consists of 4 main programs, named internal classification program, data classification program, data packet header synthesis program and kernel data processing program. The designed system is compared with the traditional system by experiments to test the working effect. The results show that the data information of bit?paralleled multi?dimension packet classification system based on cloud computing occupies a smaller space, and the classification effect is better than that of the traditional system.
Keywords: data packet classification; bit?parallelism algorithm; system design; cloud computing; bit?paralleled multi?dimension data packet; kernel data processing
0 引 言
在当代社会不断发展的大背景下,经济与科学技术也在迅猛发展,互联网技术迎来了高速发展的时代,但发展的过程必然会产生一些问题与新兴技术的挑战。目前,互联网性能的主要调节方法是提高路由器等元件的运算速度,互联网的各种新兴业务的内容查看包括数据包的路由地址查看、内部分类协议的查看以及服务器状况的查看等[1]。鉴于以上情况,互联网数据的处理就是从简单的地址查询到多维数据包分类系统的处理。路由器等基础元件存在系统紊乱的问题,而路由器的更新换代速度十分迅速,为了更好地处理数据缓冲、存储、备份问题,必须比较计算变换时间与数据查询时间[2]。……