SEED is a algorithm of Chiffrement per block developed by the agency of safety of Korean information (KISA). It is used much in the industry of the South Korea but seldom elsewhere.

SEED is a Réseau of Feistel with blocks of 128 bits and a key of 128 bits. It uses two tables of substitution of 8 X 8 which, as for SAFER, are derived from one rise to the power (in this case, x²⁴⁷ and x²⁵¹ more some “incompatible operations”). It also resembles MISTY1 in the recursivity of its structure: the coding of 128 bits is a network of Feistel with a function F operative on halves of 64 bits, while the function F itself is a network of Feistel made up of a function G working on halves of 32 bits. However recursion does not extend further because the function G is not a network of Feistel. In this function G , the word of 32 bits is regarded as 4 words of 8 bits, each one passing by one or the other of the tables of substitution, then combined via fairly complex Boolean functions so that the state of each bit of exit depends on 3 of the 4 Octet S in entry.

SEED has a complex Key schedule , generating its 32 under-keys of 32 bits through the use of the function G applying a series of rotations to the rough key, combined with constant derivative of the Golden section (as in TEA).

SEED is one of the three encryption algorithms recognized by the ISO (AES, the CAMELIA, SEED).

See too

Related articles

• MISTY1 - Network of Feistel - SAFER

External bonds

• Official document of the specifications of SEED
• RFC 4009: Encryption algorithm SEED
• RFC 4010: Use of the encryption algorithm SEED in the cryptographic syntax of message CMS