SOQUCOIN Get SOQ
Security · ML-DSA-44 FAQ

The cryptography, in the open.

Soqucoin signs every transaction with ML-DSA-44, the NIST-standardized, lattice-based signature scheme (FIPS 204). Here is exactly what that means, and the honest answers to the hard questions.

Fig. 1 · One lattice, drawn from a long, skewed basis (copper arrows) and its shortest vector (white). From a skewed basis the short vector is hard to see; at real sizes, finding it is the hard problem.

§ 01

In short

ML-DSA-44 targets NIST Security Level 2 and was published as a U.S. federal standard (FIPS 204) in August 2024, after an eight-year standardization that NIST, the NSA, GCHQ, ANSSI, and BSI all reviewed.

Soqucoin’s implementation was independently audited by Halborn.

“The most efficient secure option today, with a soft-fork path to upgrade if the threat model changes.”

Fig. 2 · A padlock, standing in for the signatures that guard coins today. Illustration.

§ 02

NIST security levels

NIST defines five post-quantum security levels, each pegged to a well-understood symmetric primitive. Level 2 is NIST’s recommended set for general-purpose digital signatures.

NIST security levels
SchemeML-DSA-44 · formerly CRYSTALS-Dilithium “Dilithium2”
NIST levelLevel 2 · recommended for digital signatures
StandardNIST FIPS 204 · August 2024
Public key / signature1,312 B / 2,420 B
Hard problemModule-LWE / Module-SIS · no known quantum advantage

Fig. 3 · One lattice, drawn from a long, skewed basis (copper arrows) and its shortest vector (white). From a skewed basis the short vector is hard to see; at real sizes, finding it is the hard problem.

§ 03

“Isn’t Level 2 the lowest?”

No. Level 1 is the lowest; Level 2 is the standard recommendation for signatures. Calling it “weak” is like calling a steel vault door weak because a bank vault exists: both exceed the threat model.

The critique usually confuses two things. The old shorthand “equivalent to AES-128” meant the total computational cost is comparable. It did not mean “run Grover’s algorithm on AES-128.” Grover’s gives only a quadratic speedup, and peer-reviewed analyses (Jaques et al. 2020; Grassl et al. 2016) show the gate depth, qubit count, and error-correction needed make such an attack impractical for the foreseeable future.

Fig. 4 · A padlock, standing in for the signatures that guard coins today. Illustration.

§ 04

Why ML-DSA-44, not 65 or 87?

Because this is a blockchain. Every transaction carries a signature and a public key, and those bytes directly govern block capacity, storage, bandwidth, and full-node sync time. ML-DSA-44’s signatures are already 2,420 bytes; the higher parameter sets multiply that cost for security beyond what the threat model requires.

And it isn’t a one-way door: Soqucoin’s consensus supports BIP9-style soft forks for cryptographic upgrades. If quantum capability advances faster than expected, or NIST revises its guidance, the chain can raise its signature level without a hard fork. Deploy the most efficient secure option today; retain the ability to upgrade.

Fig. 5 · A chain of blocks, the newest still being mined. Illustration.

§ 05

Standards & independent validation

ML-DSA was standardized through NIST’s public PQC process, with cryptographers from NIST, NSA, GCHQ, ANSSI, and BSI and hundreds of academics analyzing every known quantum attack model. NIST SP 800-227 (draft, 2025) recommends ML-DSA for new federal systems, and the DoD, financial institutions, and technology companies are migrating to FIPS 204 signatures.

Halborn independently reviewed Soqucoin’s implementation: 16,000+ lines of post-quantum code, 30 findings, all remediated (audit · case study).

Fig. 6 · A check inside a ring: a node that has verified the block for itself. Illustration.

§ 06

Primary sources

  • 01

    NIST FIPS 204

    Module-Lattice-Based Digital Signature Standard (ML-DSA), August 2024.

  • 02

    NIST PQC standardization

    The eight-year public process behind the standard.

  • 03

    Jaques et al. (2020) · Grassl et al. (2016)

    “Implementing Grover Oracles… on AES” (EUROCRYPT 2020) and “Applying Grover’s Algorithm to AES” (PQCrypto 2016). The cost analyses behind the Level-2 argument.

  • 04

    Soqucoin Whitepaper (Rev. 2)

    The full protocol and cryptographic design.

Fig. 7 · The Soqucoin mark, drawn in points.