Breaking LFSR Using Genetic Algorithm

In this paper it is shown how to find LFSR using genetic algorithm. LSFRs are part of many cryptographic structures and pseudorandom number generators. Applying genetic algorithms to Linear Feedback Shift Registers (LFSR) cryptanalysis is not quite obvious. Genetic algorithms — being one of heuristic techniques — give approximate solution. The solution could be very good, but not necessarily the best, whereas cryptographic problems require one exact answer, every other being not good enough. But as it will be shown, even if it is not intuitive, breaking LFSRs using genetic algorithms can give some interesting and promising results.