SEARCH: Symmetric Encryption Algorithm Research and Characteristic Hunting

SEARCH is a Python library that we developed to help cryptanalysts evaluate the security of block ciphers, specifically those based on substitution- permutation networks (SPN), against differential cryptanalysis attacks. The core functionality of SEARCH revolves around the capabilities of Z3 Satisfiability Modulo Theories Solver to solve various constraint systems previously modelled; these systems are created based on the parameters of the SPN, including the s-box and p-box. This modelling process implies defining a set of mathematical constraints that accurately captures the propagation of differences across the network. Through the utilization of this tool, cryptanalysts facilitate the process of identifying optimal differential characteristic, by gaining access to a suite of parametrized functionalities. Among these parameters, the fundamental ones are sbox, p-box and the number of rounds on which the differential cryptanalysis is conducted. SEARCH offers extended capabilities such as generating the optimal characteristic by adding specific constraints such as fixed input or output difference. Furthermore, the tool adds support to solve systems that find the optimal iterative differential characteristics and to refine previously identified differential characteristics to uncover the best differential for the block cipher. Nevertheless, the simplicity of using SEARCH for analyzing SPNs, even for individuals lacking programming proficiency is the differentiating factor that sets this tool apart from other available solutions. By availing SEARCH, we conducted differential cryptanalysis on three lightweight algorithms: Present, Gift-64 and Rectangle. In some scenarios, the probabilities we obtained surpass the ones recognized by scientific papers, demonstrating the consistency up to 14 rounds. This highlights SEARCH's efficacy in evaluating symmetric cryptographic algorithms' security and its reliability and robustness in replicating established results.