When DES was first published in 1977, the design criteria of its S-boxes were kept secret to avoid compromising the technique of
differential cryptanalysis (which was not yet publicly known). As a result, research in what made good S-boxes was sparse at the time. Rather, the eight S-boxes of DES were the subject of intense study for many years out of a concern that a
backdoor (a
vulnerability known only to its designers) might have been planted in the cipher. As the S-boxes are the only nonlinear part of the cipher, compromising those would compromise the entire cipher. The S-box design criteria were eventually published (in ) after the public rediscovery of differential cryptanalysis, showing that they had been carefully tuned to increase resistance against this specific attack such that it was no better than
brute force. Biham and Shamir found that even small modifications to an S-box could significantly weaken DES. Any S-box where any linear combination of output bits is produced by a
bent function of the input bits is termed a
perfect S-box. S-boxes can be analyzed using
linear cryptanalysis and
differential cryptanalysis in the form of a
Linear approximation table (LAT) or
Walsh transform and
Difference Distribution Table (DDT) or autocorrelation table and spectrum. Its strength may be summarized by the
nonlinearity (bent, almost bent) and
differential uniformity (perfectly nonlinear, almost perfectly nonlinear). ==See also==