Designing and Implementing a Robust Hybrid DNA-based Encryption Framework
Date Issued
May 2025
Author(s)
Advisor
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.
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.
File(s)![Thumbnail Image]()
Name
NOMIKOS.BSC.2025.ABSTRACT.pdf
Size
1.7 MB
Format
Adobe PDF
Checksum (MD5)
b5786026c66310df3f043c809cc43771