This is a development and community site of the Netbench Framework.
Netbench is a framework for easy evaluation and experiments with packet processing algorithms. The aim of the framework is to serve as independent platform for researchers seeking the easiest way to implement their algorithms, as well as a comparison of their algorithms with reference implementations of the other approaches.
Written in Python, Netbench currently provides framework for experiments with packet classification, pattern matching and longest prefix match (IP lookup) algorithms. In the future, authors intend to add support for experiments with packet header analysis algorithms and flow context management algorithms.
Authors of the Netbench Framework invite researchers to submit patches, new algorithms, datasets, etc. to email address netbench AT fit.vutbr.cz. After review, these patches will be added to the framework.
Netbench was presented at the ANCS 2011 conference in New York. (bibtex entry)
Packet classification algorithms added: PFCA, PFCA2.
Pattern match algorithms documentation improved.
Pattern match algorithms bugfixed, code cleaned up and improved.
Pattern match algorithms added History FA, History Counting FA, Hybrid FA, DGA.
Unit tests for pattern match significantly improved.
Various other improvements of pattern match part of the Netbench library.
Improved documentation: inludes Design document and Motivation document, as well as READMEs in most directories.
Packet classification algorithms added: HiCuts, HyperCuts, BitVector, MSPCCA.
New data sets for Longest Prefix Match
New LPM algorithms: LC Tries, Hash-Treebitmap
Includes implementations of NFA based pattern matching algorithms (BETA version).
Pattern match part bugfixed (BETA version).
Several files missing in 1.5 added.
Prefix Coloring Packet Classification Algorithm added.
Classification algorithms tests include minor bugfixes.
Uses pylibpcap library to load packets from PCAP dump.
Includes implementations of pattern matching algorithms (BETA version).
Includes implementations of longest prefix match algorithms.
Major rework of the DCFL algorithm.
Classification algorithms output polished.
Contains generated documentation of modules and classes.
First release of the Netbench Framework.
Includes implementations of three packet classification algorithms (MSCA, PHCA, DCFL).