N-Gram Symmetric Cryptography Analysis for Encrypted Message Integrity Evaluation

Abstract

Data compromise, malicious human error and unintended transfer error have been serious issues confronting message integrity. Confidential information can be sent without any fear of data alteration when there is proper utilization of effective encryption approach. The N-gram model is employed in a wide range of disciplines, including character identification, speech recognition, machine translation, and syntactic and grammatical verification. This study employed a cryptographic technique for the implementation of automating the process of N-gram analysis. Cryptographic analysis of plaintext and cypher text was conducted based on four approaches such as Unigram, Bigram, Trigram and Tetragram to determine the possible occurrence of words before and after encryption. Advanced Encryption Standard (AES) and International Data Encryption Algorithm (IDEA) were used as the symmetric cryptography algorithms. The message integrity benchmark used for iteration was 26 epochs. The frequency for each N-gram analysis approach was evaluated on unencrypted file, AES encrypted file, IDEA encrypted file. Unigram frequency analysis for file without encryption is 519, AES has129 and IDEA has 179. Digram analysis for file without encryption has frequency of 196, AES has 26 while IDEA has 34. Trigram analysis shows that file without encryption, AES and IDEA have frequency of 122, 9 and 11 respectively. Additionally, for Tetragram analysis, file without encryption, AES and IDEA have frequency of 104, 3 and 0 (that is, no iteration) accordingly. Hence, IDEA is more suitable for message integrity. The study concludes that IDEA should be used to ensure high level of security for message integrity regarding encrypted file. The study therefore, recommends that other symmetric cryptography algorithms should be considered for message integrity especially for message and data in transit.