Skip to content

The repository of 6GAN, an IPv6 multi-pattern target generation algorithm using Generative Adversarial Nets with reinforcement learning. The work has been accepted as INFOCOM 2021 accepted Paper.

Notifications You must be signed in to change notification settings

CuiTianyu961030/6GAN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

6GAN

The repository of 6GAN, an IPv6 multi-pattern target generation algorithm using Generative Adversarial Nets with reinforcement learning.

6GAN forces multiple generators to train with a multi-class discriminator and an alias detector to generate non-aliased active targets with different addressing pattern types. The rewards from the discriminator and the alias detector help supervise the address sequence decision-making process.

6GAN

The work is introduced in the 2021 IEEE International Conference on Computer Communications (INFOCOM 2021):

Tianyu Cui, Gaopeng Gou, Gang Xiong, Chang Liu, Peipei Fu, Zhen Li. 6GAN: IPv6 Multi-Pattern Target Generation via Generative Adversarial Nets with Reinforcement Learning. INFOCOM 2021: 1-10

Note: this code is based on SeqGAN and SentiGAN. Many thanks to the authors.

Requirements

Data

IPv6 Hitlist provides an IPv6 Hitlist Service to publish responsive IPv6 addresses, aliased prefixes, and non-aliased prefixes to interested researchers, which could be used as the seed addresses of target generation algorithms like 6GAN.

To use the seed dataset and the aliased prefixes data, please specify the source_file path and the aliased_prefix_file path in train.py.

Run

python train.py

Seed Classification

Overview

6GAN could freely choose one of the following three classification methods to determine the addressing patterns to be learned by generators:

  • RFC Based : Possible IPv6 addressing patterns proposed in RFC 7707.
  • Entropy Clustering : Unsupervised clustering to discover the prefix-level pattern set.
  • IPv62Vec : Pattern clusters in the vector space using context learning of address words.

Requirements

To use RFC Based and Entropy Clustering methods, you should first install the ipv6toolkit and entropy-clustering requirements respectively. Then modify the cmd codes in the two classifier class in classifier.py.

For the RFC Based method, YOUR_IPV6TOOLKIT_HOMEDIR should be the your ipv6toolkit home directory:

'cat data/source_data/responsive-addresses.work | YOUR_IPV6TOOLKIT_HOMEDIR/addr6 -i -d > data/save_data/rfc_profile.txt'

For the Entropy Clustering method, YOUR_ENTROPY_CLUSTERING_HOMEDIR is your entropy-clustering home directory:

'cat data/source_data/responsive-addresses.data | ' + \
'YOUR_ENTROPY_CLUSTERING_HOMEDIR/profiles > data/save_data/ec_profile.txt'
'cat data/save_data/ec_profile.txt | ' + \
'YOUR_ENTROPY_CLUSTERING_HOMEDIR/clusters -kmeans -k ' \
+ str(k) + ' > data/save_data/ec_cluster.txt'

Multi-Pattern Targets

When training the model, the generated targets under different epochs could be find in data\candidate_set. Each generator learns the classified data in category_data respectively and generates the pattern-specific addresses. The candidate lists could be input to Zmapv6 to obtain the active IPv6 targets.

Cite

We hope the work could help more researchers chasing IPv6 addressing patterns and target generation. If the code is helpful in your work, please cite our paper:

@inproceedings{cui20216gan,
  title={6GAN: IPv6 Multi-Pattern Target Generation via Generative Adversarial Nets with Reinforcement Learning},
  author={Cui, Tianyu and Gou, Gaopeng and Xiong, Gang and Liu, Chang and Fu, Peipei and Li, Zhen},
  booktitle={IEEE INFOCOM 2021-IEEE Conference on Computer Communications},
  pages={1--10},
  year={2021},
  organization={IEEE}
}

About

The repository of 6GAN, an IPv6 multi-pattern target generation algorithm using Generative Adversarial Nets with reinforcement learning. The work has been accepted as INFOCOM 2021 accepted Paper.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages