A NFA-based programmable regular expression match engine

  • Authors:
  • Derek Pao

  • Affiliations:
  • City University of Hong Kong

  • Venue:
  • Proceedings of the 5th ACM/IEEE Symposium on Architectures for Networking and Communications Systems
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this poster, we present a programmable hardware architecture to speed up the matching of regular expressions. The match engine is modeled as non-deterministic finite automata (NFA) with auxiliary hardware features to process repetition of sub-patterns without unrolling. The computation is table-driven and the system throughput is deterministic. The lookup tables are implemented using ternary content addressable memory (TCAM). The overall table size is approximately equal to the number of transition edges in the NFA. Incremental changes to the pattern set can be accommodated by modifying the contents of the lookup tables without reconfiguring the hardware.