Work in Progress

Cryptographic-Algorithm-Identification

Repository for Cryptographic Algorithm Identification using ML. Implements feature extraction via NIST randomness tests and ensemble learning methods (Random Forest &; Logistic Regression) to classify cryptographic algorithms. Includes Python scripts, dataset, and research references demonstrating enhanced identification accuracy. Work in progress.

Project Overview

Cryptographic algorithm identification is essential in cryptanalysis, especially when only ciphertext is available. This project extracts unique features from ciphertext using the NIST SP 800-22 randomness tests and applies ensemble learning techniques to improve identification accuracy. The current focus is on classifying algorithms like AES, 3DES, Blowfish, and RC4.

Key Features

• Feature Extraction: NIST randomness tests are employed to gather distinctive features from ciphertext.
• Modeling: Ensemble methods, including Hybrid Random Forest and Logistic Regression, are used to build a robust classification model.
• Research-Based Approach: This project builds on recent studies in cryptographic identification, ensuring that the approach is grounded in existing research.

Repository Structure

The repository is organized as follows:

Cryptographic_Algorithm_Identification/
├── data/
│   └── nist_feature_dataset.csv       # Dataset used for training and evaluation
├── docs/
│   ├── research_background/           # Folder containing research PDFs used in this project
│   │   ├── hybrid_Random_forest&logistic_regressor.pdf
│   │   ├── 3DES_8KB_2.pdf
│   │   ├── Cryptographic_Algorithm_Identification_through_Machine_Learning_for_Enhanced_Data_Security.pdf
│   │   └── Machine_Learning_for_Cryptographic_Algorithm_Ident.pdf
├── src/
│   ├── model.py                       # Main model implementation
│   ├── Dataset.py                     # Dataset handling script
│   ├── csv_creator.py                 # Script to generate CSV from data
│   └── run_nist_tests.py              # Script to run NIST randomness tests for feature extraction
├── README.md                          # Project overview and usage instructions
└── requirements.txt                   # Python dependencies