Side Channels - The application of advanced techniques for system identification to study side channels for cryptographic algorithms: cryptanalysis and design

From 01-01-2003 to 31-12-2006

Description

During the last year, researchers are starting to realize that the security of a cryptographic system does not only depend on the mathematical analysis of the algorithm itself, but also on the security of the implementation. For example, in a number of concrete implementations side channels can be identified (such as the time it takes to execute an operation the power consumption signal during the execution or electromagnetic signals). If the cryptanalyst has access to such a side channel this may render the most secure algorithm completely insecure.

The goal of this project is to combine two previously unrelated disciplines. Finding a secret key (or a secret algorithm) corresponds to an identification problem, and advanced techniques from system identification or modeling should allow to improve over existing attacks. It is clear that these attacks can be improved substantially by exploiting information in multiple side channels simultaneously, for example, time, power consumption, and electromagnetic signals observed by multiple antennas. The system can then be modeled as a MIMO (Multiple Input Multiple Output) system. The most important factors that need to be taken into account are the signal to noise ratio, heteroscadisctic variations of the measurements and the incorporation of a priori information.

The following ideas from system identification and signal processing will be explored:

1. generalized correlation-analysis;
2. system identification. System identification techniques calculate a mathematical model of the cryptographic system from input/output data.
3. advanced pattern recognition techniques, such as Least Squares Support Vector Machines (LS-SVMs), can be used to identify different instructions of a cryptographic algorithm.
4. higher order statistics and multilinear algebra;
5. Independent Component Analysis (ICA)

Team

• Bart Preneel, Co-promoter (External)
• Bart De Moor, Promoter

Financing

Funding: FWO - Research Foundation - Flanders

Program/Grant Type: FWO Research Grant - FWO Research Grant