Packet classification on multiple fields

Abstract

Routers classify packets to determine which flow they belong to, and to decide what service they should receive. Classification may, in general, be based on an arbitrary number of fields in the packet header. Performing classification quickly on an arbitrary number of fields is known to be difficult, and has poor worst-case performance. In this paper, we consider a number of classifiers taken from real networks. We find that the classifiers contain considerable structure and redundancy that can be exploited by the classification algorithm. In particular, we find that a simple multi-stage classification algorithm, called RFC (recursive flow classification), can classify 30 million packets per second in pipelined hardware, or one million packets per second in software.

Fast Parallel Network Packet Filter System based on CUDA

“…Presently, RFC algorithm 31 , which is a generalization of cross-producting 32 , is the fastest classification algorithm in terms of the worst-case performance. Bitmap compression has been used in IPv4 forwarding 33,34 and IPv6 forwarding 35 .…”

A configurable packet classification architecture for Software-Defined Networking

“…Recent research proposed improvements to packet classification efficiency based on the most popular methods, such as HyperCuts [2] or Recursive Flow Classification (RFC) [3]. RFC is of interest due to its high speed performance [4].…”

LOP: A packet classification architecture with higher throughput and lower power consumption than TCAM

“…Some of the software-based packet classification methods include linear searching [7], grid-of-tries [27], HiCuts [6], HyperCuts [21], tuple space search [25], and recursive flow classification [5]. However, none of these existing software-based packet classification methods are capable of meeting the ultra high performance requirements of modern networks [10].…”

“…5 For each configuration, the number of rules (each rule has fixed 96 bits) is varied from 32 to 1024 to measure the area. Figure 15 shows area variances with three variables: the number of stages (S), the number of segments (C) and the number of rules (Ru).…”