Cryptanalysis of speech signals ciphered by TSP using annealed Hopfield Neural Network and genetic algorithms

This paper deals with the cryptanalysis of speech signals ciphered by Time Segment Permutation. It shows that (a) if the number of segments-per-block is not large, the distance between the cepstral coefficients of the segments edges may be used as a vicinity criteria and (b) the combinatorial optimization problem of re-ordering the segments may be solved using a Neural Network improved with Simulated Annealing or a Genetic Algorithm. Experimental results are presented.