Designing and Implementing a Robust Hybrid DNA-based Encryption Framework

Abstract

This thesis introduces a hybrid encryption framework that combines Huffman encoding, AES, DNA computing with biological modifications, and chaos-based dynamic key generation driven by embedded sensor data. The proposed system targets high-security applications such as medical imaging and longterm archival storage, emphasizing robustness over speed. The encryption process begins with Huffman compression to enhance entropy and disrupt statistical patterns, followed by DNA sequence encoding enriched with biological obfuscation—such as simulated introns, codon structures, and genomic markers— to increase complexity and disguise patterns. Real-time environmental data collected via an ESP32 device dynamically modifies chaotic logistic map parameters, producing unpredictable encryption keys that evolve with ambient conditions. This integration enhances resistance to cryptanalysis, replay attacks, and static-key vulnerabilities. The decryption process ensures full reversibility despite the multi-layered transformations. Experimental validation using Shannon entropy, NPCR, UACI, chi-square tests, and the NIST SP 800-22 suite demonstrates strong randomness, diffusion, and encryption stability across varied file types and sizes. Additionally, DNA storage metrics such as base count and required physical mass are computed to assess future feasibility for synthetic DNA storage. Overall, this work establishes a new paradigm for biologically integrated, chaos-enhanced cryptography with real-time adaptability.