SILMARILS-MV: Information-Theoretically Secure Multi-Verifier Signatures with Dealer-Free Key Issuance and Linear Aggregation for Permissioned Blockchains

Introduction

Permissioned blockchains rely on digital signatures to authenticate transactions, votes, and quorum certificates. While classical signatures provide computational security, post‑quantum schemes such as lattice‑based signatures inflate message sizes and CPU usage, threatening scalability. SILMARILS‑MV (Information‑Theoretically Secure Multi‑Verifier Signatures with Dealer‑Free Key Issuance and Linear Aggregation) tackles this problem by delivering an unconditionally secure signature protocol that eliminates the need for a trusted dealer, supports unlimited hop‑by‑hop transfer, and enables linear aggregation of signatures. The result is a lightweight, quantum‑resistant authentication layer that dramatically reduces network traffic and processing load.

Core Design Principles

Unconditional Security and Multi‑Verifier Model

SILMARILS‑MV treats a signature as a vector \(\alpha = A\,e(m)\), where \(A\) is a public matrix and \(e(m)\) encodes the message. Each verifier holds a secret functional \(ci = \Gamma^T \lambda(xi)\) derived from virtual anchors \(\Gamma\) that are never revealed. This construction guarantees that even an unbounded adversary corrupting up to \(\omega\) verifiers cannot forge a valid signature with probability better than \(Q/q + 1/(q-1)\), a bound that is provably tight.

Dealer‑Free Key Issuance via MPC

Traditional information‑theoretic schemes require a trusted authority to distribute secret shares. SILMARIMS‑MV replaces this with a perfectly secure multiparty computation (MPC) protocol executed by the validator committee. For any threshold \(t < n/3\), validators jointly generate and distribute the secret functional shares without any broadcast, removing the single point of trust and simplifying deployment in decentralized environments.

Linear Aggregation and Compact Certificates

Signatures are linearly aggregatable: multiple signatures can be summed to produce a single aggregate that verifies all underlying messages. This property yields constant‑size finality certificates—only \(32(\omega+1)\) bytes—regardless of the number of validators involved. Clients receive 32‑byte signatures, and verification takes roughly 2.4 µs, making the scheme suitable for high‑throughput consensus protocols.

Performance Highlights

• Traffic Reduction: In a normal‑case analytic model with 1,000 validators, threshold notarization generates about 71 KiB of authentication traffic per validator per block—over 20× less than the ML‑DSA‑44 post‑quantum baseline.
• Throughput Gains: A Rust implementation integrated with HotStuff (100 validators) shows a 2.8× increase in transactions per second compared with ML‑DSA‑44, while consuming 67× less cryptographic CPU.
• Scalability: The issuance cost per validator becomes independent of the total committee size when the committee itself signs, enabling the protocol to scale to large validator sets without a proportional increase in overhead.

Practical Considerations for Deployment

1. Committee Setup: Validators must agree on the MPC parameters and the security parameter \(q\). The protocol tolerates up to \(\lfloor (n-1)/3 \rfloor\) Byzantine members, matching the classic BFT threshold.
2. Membership Changes: SILMARILS‑MV supports robust redistribution of secret shares, allowing validators to join or leave without re‑issuing the entire key material. Share refresh can be performed periodically to maintain forward secrecy.
3. Retroactive Enrollment: New verifiers can be enrolled after the system is live by issuing fresh functional shares derived from the existing virtual anchors, preserving the unconditional security guarantees.
4. Integration Path: Because verification is a simple matrix‑vector multiplication followed by a constant‑time check, existing blockchain clients can adopt SILMARILS‑MV with minimal code changes, swapping out the signature verification routine while retaining the surrounding consensus logic.

Conclusion

SILMARILS‑MV demonstrates that information‑theoretic security need not come at the cost of performance or operational complexity. By removing the dealer, enabling linear aggregation, and delivering microsecond‑scale verification, the scheme offers a compelling alternative to both classical and post‑quantum signatures for permissioned blockchains. Implementers seeking to future‑proof their consensus layers against quantum threats while preserving high throughput should consider SILMARILS‑MV as a viable, standards‑ready building block.