Using the y-intercept in Collision Attacks for Encryption
Mandeep Singh, Albert H. Carlson, Christopher Briscoe · 2025
Patterns are key to decrypting messages. Language is composed of repeat groupings of symbols encoded in words and the syntax of a language. Patterns bleed through encryption since all encryption algorithms can be replaced by a substitution (S) cipher. This makes it possible to analyze the patterns in cipher text using the collisions in the cipher text message. Each language has its own unique lexicon and rules, making it possible to identify both the original plain text language and the size of the block used for encryption. The ability to identify this data supports both Kerckhoffs' and Shannon's assumption that the attacker knows all of the important data about the encryption algorithm, minus the secret key. In practice, this means that an attacker can quickly focus on the data needed to prosecute an attack on an encrypted message just by analyzing collisions in the message. This paper demonstrates that such an approach is feasible and lays the foundation to explore how to derive the information needed to attack encrypted messages using data from the collisions in the cipher text.