AUTHOR: Kolokotronis Nicholas
TITLE: Non–Linear Signal Processing and Applications to Cryptography
The PhD Thesis focused on the study of synchronous stream ciphers aiming at providing confidentiality in signals or data to be stored or transmitted through open networks.
Signal processing techniques were developed for the exact evaluation and enhancement of characteristics that provide the level of security in ciphertexts. Special focus was given to pseudorandom running key generators, and to their implementation through non-linear combiners and non-linear filters.
The security of these stream ciphers strongly depends on the randomness of the pseudorandom sequence keys.
Towards this direction, our research contributed to the analysis and generation of pseudorandom sequences that simulate efficiently truly random sequences, by studying notions, like linear complexity and autocorrelation function, and introducing new ones, like sequence approximation, and higher-order statistics.