Toward Optimal-Complexity Hash-Based Asynchronous MVBA with Optimal Resilience
Pith reviewed 2026-05-23 18:58 UTC · model grok-4.3
The pith
Reducer achieves optimal O(n^2) message complexity for asynchronous MVBA while raising resilience to t < n/4 using only collision-resistant hashes.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
Reducer is an MVBA protocol that achieves O(n^2) message complexity, O(nℓ + n²λ log n) bit complexity, and O(1) time complexity while tolerating t < n/4 adaptive failures and relying exclusively on collision-resistant hash functions. Its design centers on an internal strong multi-valued Byzantine agreement (SMBA) primitive that guarantees any decided value was proposed by a correct process. Reducer++ extends the approach to t < (1/3 - ε)n for arbitrary constant ε > 0 by using random-oracle hashes to ensure termination, retaining the same asymptotic complexities with constants that depend on ε.
What carries the argument
Strong multi-valued Byzantine agreement (SMBA), a new variant of Byzantine agreement that ensures the decided value originates from a correct process and is used internally by Reducer to improve resilience without raising complexity.
If this is right
- Hash-based asynchronous MVBA can now tolerate a larger fraction of adaptive faults while preserving the optimal O(n²) message complexity achieved by prior work.
- SMBA serves as a reusable sub-protocol that enforces correct-process proposals inside larger agreement constructions.
- Reducer++ demonstrates that random-oracle modeling allows resilience to approach the classical 1/3 bound arbitrarily closely while keeping time complexity constant and bit complexity quasi-quadratic.
- The gap between optimal-complexity protocols and optimally resilient protocols is reduced, showing that collision-resistant hashes suffice for t < n/4.
Where Pith is reading between the lines
- The SMBA construction could be lifted to other primitives such as reliable broadcast or leader election that also require provenance guarantees.
- Reaching exact t < n/3 without random oracles or with standard-model hashes remains open and would close the remaining resilience gap.
- These protocols could be substituted into existing blockchain or consensus stacks to increase fault tolerance at essentially the same communication cost.
Load-bearing premise
The SMBA primitive can be realized with the claimed security and complexity bounds under only collision-resistant hashes, and this realization composes into the outer MVBA without extra assumptions on the adaptive adversary or message delivery schedule.
What would settle it
An explicit construction of SMBA (or a proof of its impossibility) that meets the stated security properties, O(n) message complexity, and near-linear bit complexity under collision-resistant hashes alone.
Figures
read the original abstract
Multi-valued validated Byzantine agreement (MVBA), a fundamental primitive of distributed computing, allows $n$ processes to agree on a valid $\ell$-bit value, despite $t$ faulty processes behaving maliciously. Among hash-based solutions for the asynchronous setting with adaptive faults, the state-of-the-art HMVBA protocol achieves optimal $O(n^2)$ message complexity, (near-)optimal $O(n \ell + n^2 \lambda \log n)$ bit complexity, and optimal $O(1)$ time complexity. However, it only tolerates $t < \frac15 n$ failures. In contrast, the best-known optimally-resilient protocol, SQ, incurs a higher bit complexity of $O(n^2 \ell + n^3 \lambda)$. This poses a fundamental question: Can a hash-based protocol be designed for the asynchronous setting with adaptive faults that simultaneously achieves optimal complexity and optimal resilience? This paper takes a significant step toward answering this question. Namely, we introduce Reducer, an MVBA protocol that retains HMVBA's optimal complexity while improving its resilience to $t < \frac14 n$. Like HMVBA and SQ, Reducer relies exclusively on collision-resistant hash functions. A key innovation in Reducer's design is its internal use of strong multi-valued Byzantine agreement (SMBA), a new variant of Byzantine agreement we introduce and construct, which ensures that the decided value was proposed by a correct process. To further advance resilience toward the optimal one-third bound, we then propose Reducer++, an MVBA protocol that tolerates up to $t < (\frac13 - \epsilon)n$ adaptive failures, for any fixed constant $\epsilon > 0$. Unlike Reducer, Reducer++ does not rely on SMBA. Instead, it employs a novel approach involving hash functions modeled as random oracles to ensure termination. Reducer++ maintains constant time complexity, quadratic message complexity, and quasi-quadratic bit complexity, with constants dependent on $\epsilon$.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript introduces Reducer, an asynchronous MVBA protocol achieving O(n²) message complexity, near-optimal O(nℓ + n²λ log n) bit complexity, O(1) time complexity, and resilience t < n/4 using only collision-resistant hash functions. The key innovation is the internal use of a new SMBA primitive that forces the decided value to originate from a correct process. It further presents Reducer++ achieving t < (1/3 - ε)n for any fixed ε > 0 via random-oracle hash functions while preserving quadratic message complexity and constant time.
Significance. If the constructions and reductions hold, the work meaningfully advances hash-based asynchronous MVBA by closing the resilience gap between HMVBA (t < n/5) and optimally resilient but higher-bit-complexity protocols such as SQ, without sacrificing the optimal message and time bounds. The explicit separation of the collision-resistant Reducer from the random-oracle Reducer++ is a clear contribution.
major comments (1)
- [Abstract] Abstract (paragraph on Reducer design): the central claim that Reducer simultaneously achieves the stated complexity bounds and t < n/4 resilience rests on the new SMBA primitive being realizable with the claimed security properties under only collision-resistant hashes and composing correctly into the outer MVBA without hidden assumptions on adaptive scheduling; the provided description does not contain the security reduction or pseudocode needed to verify this composition step.
minor comments (1)
- [Abstract] The abstract states O(1) time complexity for both protocols but does not explicitly restate the bit-complexity expression for Reducer++.
Simulated Author's Rebuttal
We thank the referee for their careful review and for acknowledging the potential significance of the work in advancing hash-based asynchronous MVBA. We address the single major comment below.
read point-by-point responses
-
Referee: [Abstract] Abstract (paragraph on Reducer design): the central claim that Reducer simultaneously achieves the stated complexity bounds and t < n/4 resilience rests on the new SMBA primitive being realizable with the claimed security properties under only collision-resistant hashes and composing correctly into the outer MVBA without hidden assumptions on adaptive scheduling; the provided description does not contain the security reduction or pseudocode needed to verify this composition step.
Authors: The full manuscript contains the requested details beyond the abstract. Section 4 defines the SMBA primitive, gives its pseudocode construction based solely on collision-resistant hash functions, and provides the security reduction establishing that any decided value must originate from a correct process. Section 5 then shows the composition into Reducer, proving that the overall protocol achieves the claimed O(n²) message complexity, near-optimal bit complexity, O(1) time, and t < n/4 resilience in the standard asynchronous model with adaptive adversaries; the proof relies only on the usual scheduling assumptions of that model and introduces no hidden conditions. We are happy to add an explicit forward reference from the abstract to these sections if the referee finds that would improve readability. revision: partial
Circularity Check
No significant circularity in derivation chain
full rationale
The paper presents new protocol constructions (Reducer using SMBA, and Reducer++) rather than mathematical derivations or fitted parameters. Central claims rest on standard assumptions (collision-resistant hashes, random oracles) and explicit new primitives introduced in the work. No self-definitional equations, fitted inputs renamed as predictions, or load-bearing self-citations appear in the abstract or described design. The argument is self-contained against external benchmarks with no reduction of results to inputs by construction.
Axiom & Free-Parameter Ledger
axioms (2)
- domain assumption Collision-resistant hash functions exist and can be used to build secure protocols against adaptive adversaries
- domain assumption Random oracle model accurately models hash functions for termination arguments in Reducer++
invented entities (1)
-
SMBA (strong multi-valued Byzantine agreement)
no independent evidence
Reference graph
Works this paper leans on
-
[1]
Abraham, I., Asharov, G., Patra, A., Stern, G.: Asynchronous agreement on a core set in constant expected time and more efficient asynchronous VSS and MPC. In: TCC (4). Lecture Notes in Computer Science, vol. 15367, pp. 451–482. Springer (2024)
work page 2024
- [2]
- [3]
- [4]
-
[5]
Bacho, R., Lenzen, C., Loss, J., Ochsenreither, S., Papachristoudis, D.: Grandline: Adaptively secure DKG and randomness beacon with (log-)quadratic communication complexity. In: CCS. pp. 941–955. ACM (2024)
work page 2024
-
[6]
Bandarupalli, A., Bhat, A., Bagchi, S., Kate, A., Reiter, M.K.: Random beacons in monte carlo: Efficient asynchronous random beacon without threshold cryptography. In: CCS. pp. 2621–2635. ACM (2024)
work page 2024
- [7]
-
[8]
Dis- tributed Comput.16(4), 249–262 (2003)
Ben-Or, M., El-Yaniv, R.: Resilient-optimal interactive consistency in constant time. Dis- tributed Comput.16(4), 249–262 (2003)
work page 2003
- [9]
-
[10]
Bhat, A., Shrestha, N., Luo, Z., Kate, A., Nayak, K.: Randpiper - reconfiguration-friendly random beacons with quadratic communication. In: CCS. pp. 3502–3524. ACM (2021)
work page 2021
-
[11]
Bracha, G.: Asynchronous byzantine agreement protocols. Inf. Comput.75(2), 130–143 (1987)
work page 1987
-
[12]
The latest gossip on BFT consensus
Buchman, E., Kwon, J., Milosevic, Z.: The latest gossip on BFT consensus. arXiv:1807.04938v3 [cs.DC] (2018), http://arxiv.org/abs/1807.04938v3
-
[13]
Cachin, C., Guerraoui, R., Rodrigues, L.E.T.: Introduction to Reliable and Secure Distributed Programming (2. ed.). Springer (2011)
work page 2011
-
[14]
Cachin, C., Kursawe, K., Petzold, F., Shoup, V.: Secure and efficient asynchronous broadcast protocols. In: CRYPTO. Lecture Notes in Computer Science, vol. 2139, pp. 524–541. Springer (2001)
work page 2001
-
[15]
Cachin, C., Kursawe, K., Shoup, V.: Random oracles in constantinople: Practical asynchronous byzantine agreement using cryptography. J. Cryptol.18(3), 219–246 (2005)
work page 2005
- [16]
- [17]
- [18]
-
[19]
Catalano, D., Fiore, D.: Vector commitments and their applications. In: Public Key Cryptog- raphy. Lecture Notes in Computer Science, vol. 7778, pp. 55–72. Springer (2013)
work page 2013
-
[20]
arXiv:2501.00214v1 [cs.CR] (2024), http://arxiv.org/abs/2501.00214v1
Chen, J.: Ociormvba: Near-optimal error-free asynchronous mvba. arXiv:2501.00214v1 [cs.CR] (2024), http://arxiv.org/abs/2501.00214v1
- [21]
-
[22]
arXiv:2002.08765v1 [cs.DC] (2020),http://arxiv.org/abs/2002.08765v1
Crain, T.: Two more algorithms for randomized signature-free asynchronous binary byzan- tine consensus with t < n/ 3 and O(n2) messages and O(1) round expected termination. arXiv:2002.08765v1 [cs.DC] (2020),http://arxiv.org/abs/2002.08765v1
-
[23]
Crain, T., Gramoli, V., Larrea, M., Raynal, M.: DBFT: efficient leaderless byzantine consensus and its application to blockchains. In: NCA. pp. 1–8. IEEE (2018)
work page 2018
-
[24]
Das, S., Duan, S., Liu, S., Momose, A., Ren, L., Shoup, V.: Asynchronous consensus without trusted setup or public-key cryptography. In: CCS. pp. 3242–3256. ACM (2024)
work page 2024
-
[25]
Das, S., Xiang, Z., Ren, L.: Asynchronous data dissemination and its applications. In: CCS. pp. 2705–2721. ACM (2021)
work page 2021
-
[26]
Cryptology ePrint Archive, Paper 2023/1196 (2023),https://eprint.iacr.org/ 2023/1196
Das, S., Xiang, Z., Tomescu, A., Spiegelman, A., Pinkas, B., Ren, L.: Verifiable secret sharing simplified. Cryptology ePrint Archive, Paper 2023/1196 (2023),https://eprint.iacr.org/ 2023/1196
work page 2023
-
[27]
Das, S., Yurek, T., Xiang, Z., Miller, A., Kokoris-Kogias, L., Ren, L.: Practical asynchronous distributed key generation. In: SP. pp. 2518–2534. IEEE (2022)
work page 2022
-
[28]
Duan, S., Wang, X., Zhang, H.: FIN: practical signature-free asynchronous common subset in constant time. In: CCS. pp. 815–829. ACM (2023)
work page 2023
-
[29]
Cryptology ePrint Archive, Paper 2024/479 (2024),https://eprint.iacr.org/2024/479
Feng, H., Lu, Z., Mai, T., Tang, Q.: Making hash-based MVBA great again. Cryptology ePrint Archive, Paper 2024/479 (2024),https://eprint.iacr.org/2024/479
work page 2024
-
[30]
Cryptology ePrint Archive, Paper 2024/1710 (2024),https://eprint.iacr.org/2024/1710
Feng, H., Lu, Z., Tang, Q.:eoptimal adaptively secure hash-based asynchronous common subset. Cryptology ePrint Archive, Paper 2024/1710 (2024),https://eprint.iacr.org/2024/1710
work page 2024
-
[31]
Fischer, M.J., Lynch, N.A., Paterson, M.: Impossibility of distributed consensus with one faulty process. J. ACM32(2), 374–382 (1985)
work page 1985
- [32]
- [33]
- [34]
- [35]
- [36]
-
[37]
Chapman and Hall/CRC Press (2007)
Katz, J., Lindell, Y.: Introduction to Modern Cryptography. Chapman and Hall/CRC Press (2007)
work page 2007
-
[38]
Kimmett, B.: Improvement and partial simulation of King & Saia’s expected-polynomial-time Byzantine agreement algorithm. Ph.D. thesis, University of Victoria, Canada (2020) 31
work page 2020
-
[39]
King, V., Saia, J.: Byzantine agreement in expected polynomial time. J. ACM63(2), 13:1– 13:21 (2016)
work page 2016
-
[40]
arXiv:1812.10169v2 [cs.DC] (2018),http://arxiv.org/abs/1812.10169v2
King, V., Saia, J.: Correction to byzantine agreement in expected polynomial time, jacm 2016. arXiv:1812.10169v2 [cs.DC] (2018),http://arxiv.org/abs/1812.10169v2
- [41]
- [42]
-
[43]
Melnyk, D.: Byzantine Agreement on Representative Input Values Over Public Channels. Ph.D. thesis, ETH Zurich, Zürich, Switzerland (2020)
work page 2020
-
[44]
Merkle, R.C.: A digital signature based on a conventional encryption function. In: CRYPTO. Lecture Notes in Computer Science, vol. 293, pp. 369–378. Springer (1987)
work page 1987
-
[45]
Mostéfaoui, A., Moumen, H., Raynal, M.: Signature-free asynchronous binary byzantine con- sensus with t < n/ 3, O(n2) messages, and O(1) expected time. J. ACM 62(4), 31:1–31:21 (2015)
work page 2015
-
[46]
Acta Informatica54(5), 501–520 (2017)
Mostéfaoui, A., Raynal, M.: Signature-free asynchronous byzantine systems: from multivalued to binary consensus witht < n/ 3, O(n2) messages, and constant time. Acta Informatica54(5), 501–520 (2017)
work page 2017
- [47]
- [48]
-
[49]
Journal of the Society for Industrial and Applied Mathematics8(2), 300–304 (1960)
Reed, I.S., Solomon, G.: Polynomial codes over certain finite fields. Journal of the Society for Industrial and Applied Mathematics8(2), 300–304 (1960). https://doi.org/10.1137/0108018
- [50]
-
[51]
Cryptology ePrint Archive, Paper 2023/1549 (2023),https://eprint
Sui, X., Wang, X., Duan, S.: Signature-free atomic broadcast with optimalO(n2) messages and O(1) expected time. Cryptology ePrint Archive, Paper 2023/1549 (2023),https://eprint. iacr.org/2023/1549
work page 2023
-
[52]
Zhang, H., Duan, S.: PACE: fully parallelizable BFT from reproposable byzantine agreement. In: CCS. pp. 3151–3164. ACM (2022) A MBAℓ: Pseudocode & Proof This section presentsMBAℓ, our adaptively-secure asynchronous MBA protocol employed by both Reducer and Reducer++. MBAℓ exchanges O(n2) messages and O(nℓ + n2λ log n) bits, and termi- nates in O(1) time. ...
work page 2022
-
[53]
Each correct process broadcasts its proposal using theCRBλ algorithm
-
[54]
Due to the justification property of the CRBλ algorithm, v was proposed by a correct process toSMBAλ
Each correct process delivers a valuev ∈ X from CRBλ. Due to the justification property of the CRBλ algorithm, v was proposed by a correct process toSMBAλ
-
[55]
(b) Decide some valuev′ ∈ X ∪ {⊥ MBA}
For eachindex = 1, 2, ..., x − 1, execute the following logic: (a) Propose v to the index-th instance of the MBA primitive. (b) Decide some valuev′ ∈ X ∪ {⊥ MBA}. Now, we differentiate two cases: • If v′ ̸= ⊥MBA, then decidev′ and terminate. Importantly, the justification property of the MBA primitive ensures thatv′ was proposed by a correct process (toSM...
-
[56]
Decide the lexicographically smallest value from theX set. If this step is reached, it means that all values from theX set have been proposed by correct processes toSMBAλ; otherwise, correct processes would have decided in one of thex − 1 iterations. As |X | = x ∈ O(1), the modifiedSMBAλ algorithm exchanges O(xn2) messages and O(xn2ℓ) bits (where ℓ denote...
-
[57]
Therefore, quality is ensured as the probability that an adversarial value is decided is at most5 6 < 1. C.2 Proof of Complexity We now formally prove the complexity ofReducer (see Thm. 2). Message complexity.We start by proving thatReducer sends O(n2) messages in expectation. Lemma 13 (Reducer’s expected message complexity). Given n = 4 t + 1 and the exi...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.