【24h】

A bit-split byte-parallel string matching architecture

机译:位分割字节并行字符串匹配体系结构

获取原文

摘要

String matching is one of hot spots in computer science. Many algorithms for string matching have been proposed, but designing an efficient and practical string matching architecture to satisfy high-speed data streaming is difficult and needs further study. In this paper we propose a high throughput configurable string matching architecture based on Aho-Corasick algorithm. The architecture can be realized by random-access memory (RAM) and basic logic elements instead of designing new dedicated chips. The bit-split technique is used to reduce the RAM size, and the byte-parallel technique is used to boost the throughput of the architecture. By the particular design and comprehensive experiments with 100MHz RAM chips, one piece of the architecture can achieve a throughput of up to 1.6Gbps by 2-byte-parallel input, and we can further boost the throughput by using multiple parallel architectures.
机译:字符串匹配是计算机科学中的热点之一。已经提出了许多用于字符串匹配的算法,但是设计一种高效且实用的字符串匹配体系结构以满足高速数据流是困难的,需要进一步研究。本文提出了一种基于Aho-Corasick算法的高吞吐量可配置字符串匹配架构。该架构可以通过随机存取存储器(RAM)和基本逻辑元件来实现,而不必设计新的专用芯片。位拆分技术用于减小RAM大小,而字节并行技术用于提高体系结构的吞吐量。通过100MHz RAM芯片的特殊设计和全面实验,一种架构可以通过2字节并行输入实现高达1.6Gbps的吞吐量,并且我们可以通过使用多种并行架构进一步提高吞吐量。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号