Stateless Post-Quantum Threshold Signatures and Knowledge-Asymmetry Frozen UTXO Rescue for Blockchain Custody and Legacy Supply Migration
Institutional cryptocurrency custodians and legacy unspent transaction outputs face two unresolved threats during the transition to post-quantum cryptography. First, institutional custody engines operate distributed signing servers across isolated availability zones and restore key material from cold backups, making stateful post-quantum signature schemes unsafe due to counter desynchronization and state reuse vulnerabilities. Second, roughly 34 percent of the Bitcoin supply resides in legacy addresses vulnerable to quantum public-key recovery or scheduled sunsetting under draft BIP-361. This paper introduces two protocol constructions built on the Adaptive Quantum Migration Protocol (AQMP v2.0). First, we present a Stateless Post-Quantum Threshold Multi-Party Computation (PQ-MPC) signature engine over module lattices (q = 8,380,417) using synthetic pseudorandom nonce derivation bound to message digests, share commitments, active quorum identifiers, and backup-safe epoch salts. Empirical benchmarks confirm zero mutable signing state across cold-backup wipe and restore events, constant-time verification between 3.90 ms and 4.68 ms across 2-of-3, 3-of-5, and 7-of-11 quorums, and a 99.71 percent block footprint reduction at N = 250 transactions (11,300 bytes via Plonky3 KeyRegistry aggregation versus 3,945,750 bytes for unaggregated 3-of-5 ML-DSA-65). Second, we propose a Knowledge-Asymmetry Frozen UTXO Rescue protocol that proves possession of unhardened BIP-32 parent chain codes and pre-compromise derivation paths in zero knowledge without exposing raw elliptic-curve public keys on-chain. Bound to a 96-byte Dual-Commit XOR cycle and the consensus confirmation deadline ceil(4 / (1 - epsilon)), the complete rescue payload requires 676 bytes, reducing on-chain witness overhead by 80.23 percent compared to Lifeboat (3,420 bytes), 82.68 percent compared to Dropkick (3,904 bytes), and 87.15 percent compared to naive ML-DSA-65 (5,261 bytes).
Authors
- Laudza Kusuma
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.5281/zenodo.24341313
- Primary Topic
- Cryptography and Data Security
- Type
- preprint