Measuring Algorithm Efficiency in Modern Password Crackers

Justin Riccardelli and Dr. Jason Pittman, Computer Science, High Point University, One University Parkway, High Point NC 27268

On average, people spend close to seven hours a day online. Activities range from accessing work systems, banking, shopping, gaming, and of course social media. Naturally, all these activities require users to authenticate in order to use those services. Authentication typically means passwords and has largely remained the same for decades. The only thing that has changed is the number of passwords a normal user must manage. That number is up to roughly 80 passwords in 2020. Without a doubt, passwords are the lifeblood of modern online existence. Still, as formidable as passwords have become, there are a myriad of attacks aimed at defeating them. The most rudimentary attack involves blind guessing. In contrast, state-of-the-art attacks use natural language processing and time-trade off vectors to optimize password cracking functions. In fact, the focus of password cracking optimization research is on speeding up hash or cracking functions and on designing customized hardware to speed up hashing or cracking functions. What does not exist is an experimental comparison of efficiency characteristics related to different programming language implementation of common password cracking algorithms. In this work, two implementations of time-trade off password cracking algorithms- Python and C languages- were run against a dataset containing publicly released passwords. Program efficiency characteristic of elapsed run time and compute resource consumption were analyzed for statistical differences and significance.

Additional Abstract Information

Presenter: Justin Riccardelli

Institution: High Point University

Type: Poster

Subject: Computer Science

Status: Approved

Time and Location

Session: Poster 5
Date/Time: Tue 12:30pm-1:30pm
Session Number: 4000