- HOME
- Electrical Technologies / Informatics
- Research Data (No.467)

## Fast Data Encipherment Algorithm (FEAL)

The problem of security is becoming increasingly important as the information network takes on a larger role in society. When speaking of “security,” consideration must first be given to achieving a measure of reliability against natural disasters and unforeseen accidents. But it is also essential that safeguards be given against malicious behavior as in computer viruses that could involve ordinary working members of society. Against this background, the need has been felt for a simple and economical cipher to provide a safe information network.

A cipher converts a document (plaintext) into a collection of symbols and/or characters that can only be decoded by the parties concerned, i.e., cannot be deciphered by a third party. In past ciphers, it was commonly accepted that both the conversion method (encryption algorithm) and conversion parameters (encryption/decryption keys) be kept secret. However, if encryption functions are to be provided to many users in an economical manner for use on a public network, a cipher of the “public-algorithm type” that can guarantee safety with only keys made secret is desirable.

A famous cipher of the public-algorithm type called DES became a U.S. standard in 1977. This algorithm, however, suffered from slow software-based encryption processing and low throughput (processing performance per unit time), and consequently, a cipher that could perform high-speed processing even by software was eagerly anticipated. Yet little theoretical research had been performed on “cipher strength” that was essential for the research of high-speed ciphers of the public-algorithm type.

It was under these conditions that two key papers titled “Evaluation Criteria for Encipherment and. Authentication Algorithms” and “Fast Data Encipherment Algorithm FEAL” were presented in 1986. Focusing on cipher input and output information, these papers proposed a cipher-evaluation theory and a method for calculating evaluation criteria. Here, the degree of independence between change in plaintext or key and change in resulting cipher text is taken to be a “strength indicator.” The validity of this “strength theory” was tested and confirmed using DES and other ciphers thereby laying the foundation for the design of a public-algorithm type of cipher with high cipher strength.

Research then began on a public-algorithm type of cipher that could perform high-speed processing despite the small scale of microprocessor software and that could perform enciphering and deciphering using the same algorithm. These efforts led to the development of a cipher called Fast Data Encipherment Algorithm (FEAL) that satisfied these requirements. This cipher, of course, satisfied cipher-strength conditions and could decipher a cipher text into the original plaintext perfectly.

In developing FEAL, its designers came up with an original “data-mixing process” whereby 64-bit blocks of plaintext and a 64-bit key are subjected to logical operations in 8-bit units as part of the enciphering process. Despite its simple structure, the effect of this mixing process exceeded that of the DES cipher achieving a sufficiently high level of security at only half the number of mixing steps (8) of DES. The FEAL cipher broke down commonly accepted beliefs attaining a cipher processing speed of over 200 kbit/s (several tens times that of DES) at a microprocessor software scale of 400 bytes at the most. It represented a landmark in the development of practical software-based ciphers.

Being a public-algorithm type of cipher, FEAL opened the way for the use of ciphers by a wide range of users. FEAL can be easily incorporated in microprocessor-based telecommunications devices because of its compact and high-speed features, and it has come to be used in smart cards and telecommunication systems.

In recognition of their research and development of the FEAL cipher, Shoji Miyaguchi, Akira Shiraishi, and Akihiro Shimizu (NTT) received the “Fifth Kobayashi Memorial Special Award” and an Achievement Award from the Institute of Electronics, Information and Communication Engineers (IEICE) in 1989.

A cipher converts a document (plaintext) into a collection of symbols and/or characters that can only be decoded by the parties concerned, i.e., cannot be deciphered by a third party. In past ciphers, it was commonly accepted that both the conversion method (encryption algorithm) and conversion parameters (encryption/decryption keys) be kept secret. However, if encryption functions are to be provided to many users in an economical manner for use on a public network, a cipher of the “public-algorithm type” that can guarantee safety with only keys made secret is desirable.

A famous cipher of the public-algorithm type called DES became a U.S. standard in 1977. This algorithm, however, suffered from slow software-based encryption processing and low throughput (processing performance per unit time), and consequently, a cipher that could perform high-speed processing even by software was eagerly anticipated. Yet little theoretical research had been performed on “cipher strength” that was essential for the research of high-speed ciphers of the public-algorithm type.

It was under these conditions that two key papers titled “Evaluation Criteria for Encipherment and. Authentication Algorithms” and “Fast Data Encipherment Algorithm FEAL” were presented in 1986. Focusing on cipher input and output information, these papers proposed a cipher-evaluation theory and a method for calculating evaluation criteria. Here, the degree of independence between change in plaintext or key and change in resulting cipher text is taken to be a “strength indicator.” The validity of this “strength theory” was tested and confirmed using DES and other ciphers thereby laying the foundation for the design of a public-algorithm type of cipher with high cipher strength.

Research then began on a public-algorithm type of cipher that could perform high-speed processing despite the small scale of microprocessor software and that could perform enciphering and deciphering using the same algorithm. These efforts led to the development of a cipher called Fast Data Encipherment Algorithm (FEAL) that satisfied these requirements. This cipher, of course, satisfied cipher-strength conditions and could decipher a cipher text into the original plaintext perfectly.

In developing FEAL, its designers came up with an original “data-mixing process” whereby 64-bit blocks of plaintext and a 64-bit key are subjected to logical operations in 8-bit units as part of the enciphering process. Despite its simple structure, the effect of this mixing process exceeded that of the DES cipher achieving a sufficiently high level of security at only half the number of mixing steps (8) of DES. The FEAL cipher broke down commonly accepted beliefs attaining a cipher processing speed of over 200 kbit/s (several tens times that of DES) at a microprocessor software scale of 400 bytes at the most. It represented a landmark in the development of practical software-based ciphers.

Being a public-algorithm type of cipher, FEAL opened the way for the use of ciphers by a wide range of users. FEAL can be easily incorporated in microprocessor-based telecommunications devices because of its compact and high-speed features, and it has come to be used in smart cards and telecommunication systems.

In recognition of their research and development of the FEAL cipher, Shoji Miyaguchi, Akira Shiraishi, and Akihiro Shimizu (NTT) received the “Fifth Kobayashi Memorial Special Award” and an Achievement Award from the Institute of Electronics, Information and Communication Engineers (IEICE) in 1989.