Encryption Paradigm Comparison Framework A comprehensive framework for comparing the efficacy and feasibility of quantum, machine learning-based, and classical encryption techniques. Overview This project implements a comparative analysis system for evaluating different encryption paradigms across multiple dimensions:

Performance: Encryption/decryption speed, throughput, and scaling with data size Security: Key strength, resistance to classical and quantum attacks Implementation complexity: Code complexity, dependencies, integration effort Resource usage: CPU utilization, memory consumption

The framework includes implementations of:

Quantum encryption: BB84 protocol, Quantum Random Number Generation (QRNG), E91 protocol ML-based encryption: GPT-based key generation, BERT-based secure protocol, TransformerXL cryptography Classical encryption: RSA, AES, and 3DES implementations

Features

Standardized testing framework for fair comparison across paradigms Comprehensive metrics collection covering performance, security, and implementation aspects Automated analysis pipeline with statistical validation Visualization tools for comparing results across dimensions Use-case specific evaluations to guide practical application

Requirements

Python 3.9+ NumPy, Matplotlib, Pandas, SciPy PyTorch (for ML-based encryption) Qiskit (for quantum encryption) Transformers library (Hugging Face) PyCryptodome (for classical encryption) scikit-learn, tqdm, tabulate

Installation bash# Create and activate virtual environment (recommended) python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate

Install required packages

pip install numpy matplotlib pandas scipy torch qiskit transformers pycryptodome scikit-learn tqdm tabulate Usage Running the Complete Comparison The main entry point is test_harness.py, which runs all tests and generates reports: bash# Run standard tests python test_harness.py --mode standard --output results

Run quick tests (smaller dataset, faster execution)

python test_harness.py --mode quick --output results

Run comprehensive tests (larger dataset, slower execution)

python test_harness.py --mode comprehensive --output results

Test specific encryption categories

python test_harness.py --subset quantum python test_harness.py --subset ml python test_harness.py --subset classical Running Individual Components Each encryption method can be tested independently: bash# Test classical encryption python classical_encryption.py

Test ML-based encryption

python ml_encryption.py

Test quantum encryption protocols

python quantum_bb84.py python quantum_qrng.py python quantum_e91.py Analyzing Results After running tests, results are available in the specified output directory:

HTML report (report/summary_report.html) Performance summary CSV (performance_summary.csv) Security summary CSV (security_summary.csv) Implementation complexity metrics (implementation_summary.csv) Visualizations (various PNG files in the report directory)

Project Structure

classical_encryption.py: RSA, AES, and DES implementations ml_encryption.py: ML-based cryptographic techniques quantum_bb84.py: BB84 quantum key distribution protocol quantum_qrng.py: Quantum random number generation quantum_e91.py: E91 quantum entanglement protocol comparative_analysis.py: Framework for comparative analysis data_collection.py: Data collection and analysis utilities test_harness.py: Main test runner and report generator

Extending the Framework To add new encryption algorithms:

Create a class implementing encrypt and decrypt methods Register the algorithm in test_harness.py Run the comparison tests to see how it compares to existing methods

Performance Notes

Quantum simulations are computationally intensive and benefit from systems with ample RAM ML-based methods perform better with GPU acceleration The comprehensive test mode can take several hours to complete; use quick mode for initial testing

Attribution This project is part of a research study "Evaluating the Efficacy and Feasibility of Quantum Encryption Versus Machine Learning-Based and Classical Encryption Techniques" by Ibrahim Al-Rawi.

Citation If you use this framework in your research, please cite: Al-Rawi, I. (2024). Evaluating the Efficacy and Feasibility of Quantum Encryption Versus Machine Learning-Based and Classical Encryption Techniques.

alrawii/Dissertation_FULLCODE

About

Resources

Stars

Watchers

Forks

Releases

Languages