Schnorr signature verification
threshold signatures
multi-sigs
blind signatures
...
on-chain
off-chain
t−1 malicious signers
cannot produce
a valid signature.
t honest signers
can reliably produce
a valid signature
(even if other n−t signers are malicious).
t-of-n
"FROST: Flexible Round-Optimized Schnorr Threshold Signatures" (Komlo and Goldberg 2020)
Produces ordinary Schnorr signatures
Problem: FROST does not provide Robustness.
ROAST is a wrapper around FROST that
turns it into a robust threshold signing protocol.
Coordinator
Coordinator not trusted
for unforgeability
Signer
Signer
Signer
Signer
2-of-4 example
2-of-4 example
2-of-4 example
2-of-4 example
?
2-of-4 example
Run a session for every subset of size t.
?
1
1
?
1
1
1
1
2
2
?
Theorem (Robustness, informal)
The coordinator outputs a valid signature after initiating at most n−t+1 signing sessions of FROST.
Proof idea.
Termination: Every signer can hold up at most one session, so n−t malicious signers can hold up at most n−t sessions.
Progress: t honest signers will respond eventually, so we can always eventually start a new session.
11-of-15 | 67-of-100 | |
---|---|---|
Uncoordinated Attack | 0.6 s | 0.7 s |
Coordinated Attack | 1 s | 7 s |
ROAST turns FROST it into a threshold signing protocol that
Paper: https://ia.cr/2022/550
Prototype: https://github.com/robot-dreams/roast