Statistical analysis and comparison of linear regression attacks on the advanced encryption standard
Hiren B. Patel, Christine M. Schubert Kabban, Rusty O. Baldwin, David P. Montminy · International Journal of Information and Communication Technology · 2015
This research investigates profiled linear regression–based attacks for extracting the advanced encryption standard (AES) secret key. Several methods from recent advancements are compared for their capability to correctly build the multivariate distribution for profiling. Attack performance shows greater than 98% success rate with as few as 100 training and test traces. In 8 out of 9 test cases examined, linear regression attacks using the coefficient of determination R2, adjusted coefficient of determination R2a and correlation power analysis (CPA) performed better than or equal to the original stochastic attack and attack using the symmetry metric. Our new method using R2a is proven to suppress unimportant variables and enhance important ones better than other methods. It is successful when the microcontrollers and data collection hardware differ between training and test phases and is found to be more effective in noisy environments than CPA.