An encryption scheme for large Chinese texts

Abstract

We improve the encryption of oriental language texts in a high-secure official document system such as military or government. As an example, an encryption scheme based on DBS is introduced for large Chinese text files. The ideogram-based Chinese texts are different from the alphabet-based English texts in many important features. First, an English text is a sequence of one-byte characters. A Chinese text is also a sequence of characters, but usually two-bytes of characters. Second, the number of distinct characters appeared in a Chinese text is much higher than that appeared in an English text. Third, the frequency distribution of the Chinese characters is very skewed. In DES, the assumption of random and uniform distribution of characters is unrealistic, especially, for Chinese texts. DES with substitution and transposition based ciphers has been subjected to attacks such as linear cryptanalysis and differential cryptanalysis. Due to the skewness of characters' distribution, Chinese texts encrypted by DES would be vulnerable for these attacks. This paper bases on DES to design a new scheme for encrypting large Chinese texts. We first divide each two-bytes character in a Chinese text into the high-byte group and the low-byte group. They include the high-byte and the low-byte of each character, respectively. According to the distributions of the high-byte and the low-byte of Chinese characters, we recode each byte in these two groups. Finally, we encrypt the recoded bytes in these two groups by using a modified DES algorithm individually. By decrypting and combining these two groups of encrypted data, the original Chinese text can be revealed.