Protocol Learning and Stateful Fuzzing
Python Other
Latest commit baabdcc Sep 12, 2016 @hgascon committed on GitHub Update requirements.txt
Failed to load latest commit information.
models initial commit Apr 28, 2015
pulsar minimal style fix Aug 5, 2016
rsc/icon new icon Feb 24, 2016
.gitignore initial commit Apr 28, 2015
AUTHORS initial commit Apr 28, 2015
LICENSE initial commit Apr 28, 2015 Add 'chat on gitter' badge May 18, 2016 Add Simulator class May 10, 2016
requirements.txt Update requirements.txt Sep 12, 2016


Join the chat at

Protocol Learning, Simulation and Stateful Fuzzer

Pulsar is a network fuzzer with automatic protocol learning and simulation capabilites. The tool allows to model a protocol through machine learning techniques, such as clustering and hidden Markov models. These models can be used to simulate communication between Pulsar and a real client or server thanks to semantically correct messages which, in combination with a series of fuzzing primitives, allow to test the implementation of an unknown protocol for errors in deeper states of its protocol state machine.

For detailed information about the method implemented by Pulsar, you can read the following publications:

Pulsar: Stateful Black-Box Fuzzing of Proprietary Network Protocols
Hugo Gascon, Christian Wressnegger, Fabian Yamaguchi, Daniel Arp and Konrad Rieck
Proc. of 11th EAI International Conference on Security and Privacy in Communication Networks (SECURECOMM) October 2015

Learning Stateful Models for Network Honeypots
Tammo Krueger, Hugo Gascon, Nicole Krämer and Konrad Rieck
ACM Workshop on Security and Artificial Intelligence (AISEC) October 2012

     _ __  _   _| |___  __ _ _ __
    | '_ \| | | | / __|/ _` | '__|
    | |_) | |_| | \__ \ (_| | |
    | .__/ \__,_|_|___/\__,_|_|  v0.1-dev

usage: [-h] [-c CONF] [-l] [-p PCAP] [-b BINARIES] [-a] [-x]
                 [-o OUT] [-d DIMENSION] [-s] [-z] [-m MODEL]

Protocol Learning and Stateful Fuzzing

optional arguments:
  -h, --help            show this help message and exit
  -c CONF, --conf CONF  Change default directory for configuration files. If
                        no directory is given, the files from 'pulsar/conf'
                        will be read.

  -l, --learner         Learn a model from a set of network traces.
  -p PCAP, --pcap PCAP  tcpdump output file (pcap) or list of files separated
                        by commas to use as input data for a new model.
  -b BINARIES, --binaries BINARIES
                        Name of binaries to process from the cuckoo storage
                        dir separated with commas.
  -a, --all-binaries    Generate models for all binaries from the cuckoo
                        storage dir (cuckoo/storage/binaries).
  -x, --process         Process derrick files through the functions defined in
  -o OUT, --out OUT     Change output directory for generated models. If no
                        directory is given, the model will be written to the
                        'models' directory.
  -d DIMENSION, --dimension DIMENSION
                        Number of components to be used for NMF clustering.

  -s, --simulate        Simulate communication based on a given model.
  -z, --fuzzer          Start a fuzzing session based on a given model.
  -m MODEL, --model MODEL
                        Path of the dir containing the model files to be
                        loaded for simulation or fuzzing.


The directory pulsar/conf contains a series of configuration files that define the parameters required for certain operations in each one of the Pulsar methods for automatic learning, simulation and fuzzing.


Generate the model of a communication channel from individual PCAP files or the recorded traces of one or more binaries run by cuckoo sandbox:

$> -l -p file.pcap (1 pcap file)
$> -b 016169EBEBF1CEC2AAD6C7F0D0EE9026 (1 or more binaries from cuckoo storage)
$> -a (all binaries from cuckoo storage)

Simulate a communication channel based on a learnt model:

$> -s -m model_file

Initiate a fuzzing session against a target given the model of its communication channel:

$> -z -m model_file