💻 computer science

Wonderboom -- Efficient, and Censorship-Resilient Signature Aggregation for Million Scale Consensus

This paper introduces Wonderboom, a novel signature aggregation protocol that enables Ethereum to finalize blocks 32 times faster by efficiently aggregating over 2 million validator signatures within a single slot while simultaneously enhancing security against adversarial attacks and censorship.

Original authors: Zeta Avarikioti, Ray Neiheiser, Krzysztof Pietrzak, Michelle X. Yeo

Published 2026-05-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zeta Avarikioti, Ray Neiheiser, Krzysztof Pietrzak, Michelle X. Yeo

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Problem: A Traffic Jam at the World's Biggest Bank

Imagine Ethereum as a massive, global bank that holds over $650 billion in people's savings. It's so big and secure that it's more valuable than the stock exchanges in Singapore, Mexico, and Thailand combined.

To keep this bank safe, it relies on 1 million independent guards (called "validators"). These guards all have to sign off on every transaction to make it official.

The Current Bottleneck:
Right now, getting all 1 million guards to sign a single document takes forever.

The Process: The guards are split into small groups. Each group signs, then passes their signature to a bigger group, which passes it to an even bigger group, like a game of "telephone."
The Result: Because there are so many people and the "telephone" game takes time (due to internet delays), it takes about 15 minutes for a transaction to be considered "final" (safe and unchangeable).
The Risk: In those 15 minutes, a bad actor (a hacker or a greedy group) can try to block honest guards from signing. If they succeed often enough, they can steal rewards meant for honest guards, slowly taking over the bank's control.

The Solution: WONDERBOOM

The authors of this paper built a new system called WONDERBOOM. Think of it as a super-efficient, high-speed assembly line that can get all 1 million guards to sign a document in under 15 seconds (specifically, two "slots" of time), while making it much harder for bad actors to cheat.

Here is how it works, using simple metaphors:

1. The "Two-Person" Guard System (Separating Roles)

In the current system, every guard wears two hats: they vote on transactions and they announce the results. This makes it easy for hackers to figure out who is who and attack them.

WONDERBOOM's Fix:
Imagine splitting the guard into two people:

The Voter: Stays inside a secure, private room (using a direct phone line) to do the voting. No one outside can see them.
The Messenger: Stands outside in the public square (the "gossip network") to shout the final results.
Because the Messenger rarely talks to the public, it's very hard for hackers to find and attack them. This keeps the voting process fast and private.

2. The "Tree" Instead of a "Line"

Currently, signatures move in a slow, deep chain (like a long line of people passing a bucket of water). If one person drops the bucket, the whole line stops.

WONDERBOOM's Fix:
They organize the 1 million guards into a giant tree structure.

Instead of passing a message one by one, small groups of guards pass their signatures up to a "branch," which passes it to a "trunk," which passes it to the "top."
Because the tree is wide and shallow, the message travels up much faster.
The Magic Trick: To stop bad actors from blocking the "biggest" group of signatures, the system randomly picks a second group to include as well. It's like saying, "We will accept the largest pile of signatures, but we'll also randomly grab a handful from a different pile." This makes it nearly impossible for a bad actor to block everyone at once.

3. The "Reward Window" (Stopping Cheaters)

In the old system, if a guard missed their turn to sign just once, they lost their reward. This made the system very sensitive to censorship.

WONDERBOOM's Fix:
They changed the rules to a "Weekly Pass" system.

Instead of needing to sign every single day, a guard just needs to sign at least once within a 64-day window to get paid for that whole month.
Even if a bad actor blocks a guard for a few days, the guard can still get their reward as long as they slip in a signature before the window closes.
This makes it mathematically impossible for bad actors to slowly steal all the money by blocking people, because the "window" is so wide that they can't block everyone forever.

What the Paper Proved

The authors didn't just write a theory; they built a simulator (a video game of the blockchain) to test this with 1 million to 2.5 million fake guards.

Speed: They showed that WONDERBOOM can process all 1 million signatures in 4 seconds (on standard computer hardware), compared to the current 15-minute wait. That is 32 times faster.
Security: They proved that even if 1/3 of the guards are bad actors trying to censor the system, WONDERBOOM is much harder to break than the current Ethereum system.
Scalability: It works even if the number of guards grows to 2 million, without needing super-computers.

The Bottom Line

WONDERBOOM is a new way to organize a massive crowd of people so they can agree on a decision almost instantly. It solves the "traffic jam" problem of Ethereum by:

Hiding the voters to protect them from attacks.
Using a tree structure to move information faster.
Giving people a wider time window to earn rewards, making it hard for bad actors to cheat the system.

The paper concludes that we can have a blockchain that is both massively decentralized (millions of users) and instantly fast, without sacrificing security.

Technical Summary: WONDERBOOM

Problem Statement

Ethereum has evolved into a critical global financial infrastructure securing over $650 billion in assets, yet its current consensus mechanism faces significant scalability and security bottlenecks due to its validator set size (approaching one million). The primary limitations identified are:

Slow Finality: The protocol currently requires approximately 15 minutes to finalize a block. This latency stems from the high computational and bandwidth costs of aggregating and verifying signatures across millions of validators. The current design partitions validators into committees and utilizes a hierarchical tree aggregation protocol that spans multiple epochs (32 slots per epoch), requiring two consecutive epochs for finality.
Censorship Vulnerabilities: The existing aggregation protocol is susceptible to censorship attacks that can shift stake distribution from honest to adversarial nodes.
- Proposer Censorship: Adversaries can deanonymize proposers via the gossip network and censor their blocks, causing reward losses for honest validators.
- Vote Censorship: The current protocol relies on committee representatives. The paper argues that the probability of a committee having no honest representatives (all 16 selected representatives being Byzantine) is significantly higher than previously calculated metrics suggested. This allows adversaries to censor votes, leading to aggregate reward losses estimated at over $100 million annually.
- Inefficient Aggregation: The reliance on the gossip network for message propagation introduces inherent latency (approx. 4 seconds to reach 98% of nodes), limiting the depth of aggregation trees that can be constructed within a single 12-second slot.

Methodology

The authors propose WONDERBOOM, a tree-based signature aggregation protocol designed to aggregate signatures from all $N$ validators within a single Ethereum slot (12 seconds), enabling two-slot finality (approx. 24 seconds). The methodology relies on three core architectural and algorithmic innovations:

1. Architectural Separation (Validator vs. Proposer)

To mitigate proposer censorship and deanonymization, WONDERBOOM physically separates the Validator process from the Proposer process, assigning them distinct IP addresses.

Validator Process: Communicates via point-to-point (P2P) channels with other validators for aggregation. This bypasses the high-latency gossip network for the bulk of consensus traffic.
Proposer Process: Interacts with the gossip network only for block broadcasting and receiving final aggregates. The infrequency of this interaction makes deanonymization significantly harder.

2. Deep Tree Aggregation with Randomized Selection

WONDERBOOM constructs a deep aggregation tree where every node represents a committee of 128 validators (with 16 random representatives).

Structure: Unlike Ethereum's shallow tree, WONDERBOOM aggregates all $N$ signatures every slot.
Censorship Mitigation Rule: To prevent malicious aggregators from always forwarding the largest (potentially censored) aggregate, the protocol mandates that the largest aggregate be combined with one uniformly random aggregate from the child committees. This ensures that even if a malicious node controls the largest aggregate, honest votes have a non-zero probability of being included.

3. Reward Allocation and Censorship Resilience

The paper introduces a formal definition of Vote-Censorship Resilience ( $\alpha$ -resilience), defined as the probability that a validator is not censored for $k$ consecutive slots.

Windowed Rewards: Validators earn rewards for a $k$ -slot window if their vote is included at least once within that window. This lowers the per-slot inclusion probability requirement while preventing long-term stake drift.
Public Key Optimization: Leveraging the high participation rate of Ethereum (>99%), the protocol caches the aggregate of all public keys and uses BLS signature subtraction to remove missing keys, reducing public key aggregation costs by up to 99%.

Key Contributions

WONDERBOOM Protocol: The first aggregation protocol capable of efficiently aggregating millions of signatures in a single slot while offering stronger censorship resilience guarantees than the state-of-the-art Ethereum protocol.
Theoretical Analysis of Censorship: A rigorous analysis showing that the current Ethereum protocol's vote-censorship resilience is overestimated. The paper demonstrates that WONDERBOOM achieves comparable or superior resilience with a much smaller security parameter $k$ (e.g., $k=64$ in WONDERBOOM vs. $k=192$ required for Ethereum to reach similar resilience).
Simulation Framework: The development of the first simulation tool capable of modeling Byzantine fault-tolerant signature aggregation at the million-validator scale, including worst-case scenarios (e.g., $N/3$ faulty nodes).
Performance Evaluation: Empirical evidence showing that WONDERBOOM can aggregate signatures for up to 2.5 million nodes within a single slot on legacy hardware, provided point-to-point communication is used.

Results

The authors evaluated WONDERBOOM using their simulation tool on a system emulating up to 2.5 million nodes with 100ms network latency.

Latency and Scalability:
- For the current Ethereum validator set (~1 million), WONDERBOOM aggregates all signatures in under 4 seconds using only 4 CPU cores.
- For 2 million nodes, aggregation remains under 4 seconds with 8 cores.
- Even at 2.5 million nodes, the protocol remains feasible, though slightly exceeding the 4-second budget in worst-case scenarios.
Comparison with Ethereum:
- WONDERBOOM is approximately 3x faster than the current Ethereum aggregation protocol when both utilize point-to-point channels.
- When accounting for the current Ethereum reliance on the gossip network (which adds ~4 seconds of latency), WONDERBOOM is >32x faster in practice.
Censorship Resilience:
- Under a slowly adaptive adversary, WONDERBOOM achieves a 99.8% probability that no node is consistently censored over two epochs ( $k=64$ ).
- In contrast, the current Ethereum design would only guarantee an 88.9% probability under the same conditions, requiring a 6-epoch window to match WONDERBOOM's resilience.

Significance

The paper claims that WONDERBOOM demonstrates that fast finality and strong inclusion guarantees are compatible with million-validator consensus.

Decentralization Preserved: Unlike approaches that reduce the validator set or use super-committees (which compromise economic security), WONDERBOOM aggregates signatures from all validators, maintaining the full economic security of the network.
Economic Security: By mitigating vote censorship, the protocol prevents the systematic redistribution of stake from honest to adversarial nodes, a critical factor for long-term network safety.
Practical Deployability: The protocol is designed to be modular and can be integrated into existing systems like Ethereum with minimal changes, utilizing standard cryptographic primitives (BLS signatures) and existing hardware capabilities.

The authors conclude that while fully adaptive adversaries remain a theoretical challenge for single-proposer protocols, WONDERBOOM provides a robust solution against slowly adaptive adversaries, addressing the primary bottlenecks preventing Ethereum from achieving single-slot finality at scale.