Parallel Computing for Brute-Force Password Cracking

Overview

This project explores parallel computing techniques to optimize brute-force password cracking. The implementation includes two approaches: a nested loop method and a recursive algorithm, both optimized with OpenMP parallelization to enhance performance.

Implementation Details

1. Brute-Force Password Cracking

   • Simulates a brute-force attack by generating all possible character combinations.
   • Tests passwords of lengths between 2 to 6 characters.

2. Nested Loop Method (First Approach)

   • Uses multiple nested loops to generate passwords sequentially.
   • Parallelized within the main function for optimal performance.

3. Recursive Method (Second Approach)
   • Utilizes recursion to build possible password combinations dynamically.
   • Parallelized within the main function using OpenMP.

Parallelization Strategy

• Explored three parallelization methods:
  1. Parallelizing only within functions.
  2. Parallelizing only in the main function (best performance).
  3. Parallelizing both main and functions.
• The most efficient approach was parallelizing the main function.

Performance Comparison

Results & Findings

• The nested loop method proved to be the fastest for both sequential and parallel execution.
• Parallelization using OpenMP significantly reduced execution time, achieving up to 8x speedup.
• Recursive methods performed slower but demonstrated flexibility in cracking varying password lengths.

Future Improvements

• Extend the algorithm to support longer passwords with GPU acceleration.
• Implement optimized hash-based brute-force techniques.
• Explore hybrid approaches combining recursion and loops for better efficiency.

---

Contributors

• Imad-Eddine NACIRI