In cryptography, the International Data Encryption Algorithm, originally called Improved Proposed Encryption Standard, is a symmetric-keyblock cipher designed by James Massey of ETH Zurich and Xuejia Lai and was first described in 1991. The algorithm was intended as a replacement for the Data Encryption Standard. IDEA is a minor revision of an earlier cipher Proposed Encryption Standard. The cipher was designed under a research contract with the Hasler Foundation, which became part of Ascom-Tech AG. The cipher was patented in a number of countries but was freely available for non-commercial use. The name "IDEA" is also a trademark. The last patents expired in 2012, and IDEA is now patent-free and thus completely free for all uses. IDEA was used in Pretty Good Privacy v2.0 and was incorporated after the original cipher used in v1.0, BassOmatic, was found to be insecure. IDEA is an optional algorithm in the OpenPGP standard.
Operation
IDEA operates on 64-bit blocks using a 128-bit key and consists of a series of 8 identical transformations and an output transformation. The processes for encryption and decryption are similar. IDEA derives much of its security by interleaving operations from different groups — modular addition and multiplication, and bitwise eXclusive OR — which are algebraically "incompatible" in some sense. In more detail, these operators, which all deal with 16-bit quantities, are:
Multiplication modulo 216 + 1, where the all-zero word in inputs is interpreted as 216, and 216 in output is interpreted as the all-zero word .
After the 8 rounds comes a final “half-round”, the output transformation illustrated below :
Structure
The overall structure of IDEA follows the Lai–Massey scheme. XOR is used for both subtraction and addition. IDEA uses a key-dependent half-round function. To work with 16-bit words, IDEA uses the Lai–Massey scheme twice in parallel, with the two parallel round functions being interwoven with each other. To ensure sufficient diffusion, two of the sub-blocks are swapped after each round.
Key schedule
Each round uses 6 16-bit sub-keys, while the half-round uses 4, a total of 52 for 8.5 rounds. The first 8 sub-keys are extracted directly from the key, with K1 from the first round being the lower 16 bits; further groups of 8 keys are created by rotating the main key left 25 bits between each group of 8. This means that it is rotated less than once per round, on average, for a total of 6 rotations.
Decryption
Decryption works like encryption, but the order of the round keys is inverted, and the subkeys for the odd rounds are inversed. For instance, the values of subkeys K1–K4 are replaced by the inverse of K49–K52 for the respective group operation, K5 and K6 of each group should be replaced by K47 and K48 for decryption.
Security
The designers analysed IDEA to measure its strength against differential cryptanalysis and concluded that it is immune under certain assumptions. No successful linear or algebraic weaknesses have been reported. , the best attack applied to all keys could break IDEA reduced to 6 rounds. Note that a "break" is any attack that requires less than 2128 operations; the 6-round attack requires 264 known plaintexts and 2126.8 operations. Bruce Schneier thought highly of IDEA in 1996, writing: "In my opinion, it is the best and most secure block algorithm available to the public at this time." However, by 1999 he was no longer recommending IDEA due to the availability of faster algorithms, some progress in its cryptanalysis, and the issue of patents. In 2011 full 8.5-round IDEA was broken using a meet-in-the-middle attack. Independently in 2012, full 8.5-round IDEA was broken using a narrow-bicliques attack, with a reduction of cryptographic strength of about 2 bits, similar to the effect of the previous bicliques attack on AES; however, this attack does not threaten the security of IDEA in practice.
Weak keys
The very simple key schedule makes IDEA subject to a class of weak keys; some keys containing a large number of 0 bits produce weak encryption. These are of little concern in practice, being sufficiently rare that they are unnecessary to avoid explicitly when generating keys randomly. A simple fix was proposed: XORing each subkey with a 16-bit constant, such as 0x0DAE. Larger classes of weak keys were found in 2002. This is still of negligible probability to be a concern to a randomly chosen key, and some of the problems are fixed by the constant XOR proposed earlier, but the paper is not certain if all of them are. A more comprehensive redesign of the IDEA key schedule may be desirable.
Xuejia Lai and James L. Massey, , EUROCRYPT 1990, pp. 389-404
Xuejia Lai and James L. Massey and S. Murphy, Markov ciphers and differential cryptanalysis, Advances in Cryptology — Eurocrypt '91, Springer-Verlag, pp. 17-38.
