Improvement of power analysis attacks using Kalman filter

Power analysis attacks are non intrusive and easily mounted. As a consequence, there is a growing interest in efficient implementation of these attacks against block cipher algorithms such as Data Encryption Standard (DES) and Advanced Encryption Standard (AES). In our paper we propose a new technique based on the Kalman theory. We show how this technique could be useful for the cryptographic domain by making power analysis attacks faster. Moreover we prove that the Kalman filter is more powerful than the High Order Statistics technique.