Dynamically adjustable computation pattern encryption technique: boosting nonlinearity via integrating block chaotic encoding and color-theory-based key expansion

Ever-growing effectiveness of hacking techniques is making information protection more challenging nowadays than anytime ever. This state dictates devising more powerful techniques that can efficiently resist the ever-growing power of cryptoanalysis tools. Although many effective encryption techniques are available, the principal problem with these techniques is their great reliance on the complexity of the computation as a means of information protection and never using the full potential of the input space (key and plaintext). This paper proposed a technique that combines powerful computations such as dual-pass bit mixing, chaotic bit manipulation, and dynamic computation patterns. The technique has a dynamic behavior that observes the variations of the current input block and a long memory to memorize the variations of the previous blocks and adjusts its computational pattern on the fly (during the processing). This input-driven and dynamic computation pattern switching highly complicate the relation of plaintext to ciphertext. Because the proposed technique can adjust its computational state based on the inputs, it can substantially increase the nonlinearity and confusion of the ciphertext since any variation imposes a different pattern of computations. Additionally, the paper proposes a novel method based on important principles inspired by the Color theory for expanding the key to any length and highly preserving the identity of the key.