RNN Tester

This is the implementation base on the content of the paper
Testing Cryptographically Secure Pseudo Random Number Generators with Artificial Neural Networks.
The python3 program rnn_tester evaluates (cryptographically secure) pseudo random generators with AI.

Build

To build the project the python package build is required

python3 -m pip install --upgrade build

To build a package use

python3 -m build

Install

The tool requires the programm dieharder for statistical analysis of the results. Install the tool first.

• Debian/Ubuntu:

  sudo apt-get install dieharder

• Fedora:

  sudo dnf install dieharder

Install the tool with

pip install ./dist/rnn_tester-<version>-py3-none-any.whl

Usage

The tool rnn_tester requires a binary file containing the result of an PRNG in <path-to-rng-file>. Use

python3 rnn_tester [options] -i <path-to-rng-file>

to test your random numbers.

Testing

The folder test contains all unit test. All tests including statistical code analysis could be execute with

tox -r