Synchronized DNA sources for unconditionally secure cryptography

1. Problem Statement

Modern cryptography relies heavily on computational security, where encryption strength depends on the assumption that attackers lack the computational power to break algorithms (e.g., RSA, AES). This creates a fundamental vulnerability: as computing power increases (or with the advent of quantum computing), previously secure data becomes vulnerable.

While Quantum Key Distribution (QKD) offers information-theoretic security based on physical laws, it faces significant practical limitations:

• Distance: Current fiber-based QKD is limited to ~1,000 km without repeaters.
• Scalability: Satellite-based QKD is constrained by weather and limited orbital windows, resulting in low throughput (approx. 10³ kbit/day).
• Infrastructure: It requires specialized, expensive hardware and dedicated quantum channels.

The One-Time Pad (OTP) is the only known encryption method offering unconditional (perfect) secrecy, provided the key is:

1. As long as the message.
2. Truly random.
3. Used only once.
4. Shared securely between parties.

The Core Challenge: Generating and securely distributing massive, synchronized, high-quality random keys over long distances without relying on computational assumptions or quantum channels remains an unsolved problem.

2. Methodology

The authors propose a novel cryptographic primitive using synthetic DNA as a synchronized entropy source. The protocol leverages the physical properties of DNA (high density, stability, and discrete molecular nature) to generate shared secrets between distant parties (demonstrated between Paris and Tokyo).

A. The DNA-Based OTP Protocol

1. Synthesis of Random Pools:
   • Alice orders two pools of partially random DNA oligonucleotides: Index strands and Payload strands.
   • These strands contain random nucleotide blocks (A, C, G, T) separated by fixed spacers.
2. Assembly of Double-Stranded Keys:
   • The strands are annealed and extended by a polymerase to form double-stranded DNA keys (Index-Payload duplexes).
   • This creates a massive pool of unique molecular pairs.
3. Partitioning and Distribution:
   • A specific aliquot of the pool (containing ~10⁶–10⁹ molecules) is physically split into two "pads."
   • One pad is kept by Alice; the other is shipped to Bob.
   • Crucially: The information exists in exactly two copies (one direct, one reverse-complement) within the pool before splitting.
4. Synchronization and Key Generation:
   • Sequencing: Both parties independently sequence their DNA pads using nanopore technology.
   • Index Exchange: Bob publicly sends his list of Index sequences to Alice. He keeps the Payload sequences secret.
   • Authentication & Sifting: Alice compares Bob's indices with her own. The intersection of indices serves as the shared secret. The corresponding Payloads are ordered to form a binary mask.
   • Encryption: The binary mask is used as the OTP key to encrypt/decrypt messages via XOR.

B. Randomness Enhancement (5PPD)

Raw DNA synthesis often suffers from nucleotide biases and correlations. To ensure cryptographic-grade randomness:

• The authors employ Block-5 Purine Parity Digitization (5PPD).
• This method counts the number of purines (A or G) in blocks of 5 random bases modulo 2.
• This process eliminates position-dependent biases and spatial correlations, converting the DNA sequence into a high-quality binary stream.

C. Security Against Eavesdropping (UMI Tagging)

To detect interception (Eve), the protocol utilizes Unique Molecular Identifiers (UMIs):

• Scenario 1 (Partial Theft): If Eve steals a fraction of Bob's pad, the statistical overlap of the three parties (Alice, Bob, Eve) becomes null.
• Scenario 2 (Copy-and-Replace): If Eve steals the pad, amplifies it (PCR), and replaces Bob's sample, the molecular copy-number statistics change.
  • In the original protocol, a key exists as a single pair (1 direct + 1 reverse complement).
  • Eve's amplification creates multiple copies.
  • By counting UMIs per cluster, Alice and Bob can detect anomalies in the distribution of copy numbers (e.g., clusters with >2 UMIs), revealing the attack.

3. Key Contributions

• First DNA-Based OTP Implementation: Demonstrates a working OTP system using synthetic DNA as a shared entropy source between two continents (Paris and Tokyo).
• Unconditional Security without Quantum Limits: Provides a scalable alternative to QKD that is not limited by distance, weather, or the need for quantum repeaters.
• High Entropy Generation: The generated binary masks meet NIST SP 800-90B standards for min-entropy, comparable to approved hardware random number generators.
• Intrusion Detection: Introduces a method to detect eavesdropping based on molecular copy-number statistics (UMI analysis), adding a layer of physical security reminiscent of QKD but applicable to classical molecular systems.
• Scalability: Theoretically capable of generating petabytes of keys from a single gram of DNA, with installation occurring only once.

4. Experimental Results

• Throughput: Generated a shared secret mask of ~400 Mb (specifically 316 Mb after filtering) in a single run.
• Error Rate: The residual error rate between Alice's and Bob's masks was ~5 × 10⁻⁵.
• Correction: Using Bose–Chaudhuri–Hocquenghem (BCH) error-correcting codes, the system achieved a decryption failure rate of 2⁻¹²⁸, meeting standard cryptographic security requirements.
• Randomness Quality:
  • Min-entropy values: 0.9588 (Alice) and 0.9604 (Bob).
  • These values are on par with NIST-approved deterministic RNGs (e.g., Hash-DRBG-SHA256).
• Demonstration: Successfully encrypted, transmitted, and decrypted a 130 Mb color image (Horsehead Nebula) between the two locations.
• Attack Simulation: Successfully simulated and detected two types of adversarial attacks (partial theft and copy-and-replace) using UMI-based statistical analysis.

5. Significance and Future Outlook

This work bridges molecular biology and cryptography, establishing DNA as a viable, scalable medium for unconditional security.

• Strategic Advantage: Unlike QKD, DNA pads can be stored for centuries and transported physically with minimal frequency. Once installed, key generation requires no transfer of confidential data, only public index exchange.
• Applications: Ideal for high-security, long-term archival, and sensitive communications (finance, defense, elections) where computational security is insufficient.
• Challenges: Current limitations include latency and the cost of sequencing. However, the authors note that sequencing costs are dropping exponentially, and automation could make DNA-OTP accessible for high-bandwidth applications in the near future.
• Conclusion: The paper proves that molecular randomness can be harnessed to create a "Holy Grail" of cryptography: a synchronized, unclonable, and perfectly random key source that operates independently of distance and computational power.