Braid: Enhancing Ethereum’s Censorship Resistance Through Technological Innovation

As blockchain technology evolves, censorship resistance has become a key concern in the industry. Since Ethereum’s merge upgrade, the sector has actively sought solutions for effective censorship resistance under the new consensus mechanism. In this context, Braid has emerged as a promising solution. Vitalik’s latest comments mentioned that “the debate mainly centers around FOCIL+APS versus BRAID,” accompanied by graphs showing the scale of Ethereum stakers and BTC mining pool distributions.

Ethereum staker distribution and BTC mining pool share chart (Source: X)

Braid, proposed by Max Resnick at Paradigm’s New York workshop, is a method for implementing multi-block parallelism on Ethereum aimed at building a censorship-resistant system. This article will explore Braid’s background, technical features, and comparisons with other existing solutions to provide a more professional and in-depth understanding, helping readers gain a multi-faceted understanding of the Ethereum Braid project.

Background of Braid

Both Bitcoin and Ethereum face a common issue: the tendency of mining pools and validation nodes to centralize. This centralization could lead to an attack in which mining pools or validation nodes might be legally required to follow regulations, blocking or censoring transactions deemed illegal within their control.

The graph shows that Lido and Coinbase collectively hold a 40.8% share. If these two were to unite, they could potentially halt the network. If over 10% more staking service providers join them, they could take over the Ethereum network. Before discussing the centralization threats to nodes, let’s examine how Ethereum operates under the PoS mechanism.

Transactions submitted by users through dapp frontends enter the Mempool, a database of pending transactions. At this point, Searchers use arbitrage bots to analyze the backlogged transactions in the mempool, combining profitable transactions into transaction groups or bundles. Block builders receive these bundled transactions from searchers, attach corresponding processing fees to these bundles, and prepare bids. Validators (block proposers) select the most profitable option from the provided bundles and bids, then formulate and propose new blocks. It undergoes final checks and validation before a block is added to the chain.

MEV Supply Chain（Source: FlashBot）

This mechanism often leads to high concentrations of “validators” or potential collusion for malicious purposes, increasing the risk of network censorship and control. Braid emerged in response to these challenges, aiming to create an Ethereum framework resistant to censorship. Currently in its early stages, Braid’s main idea is to break the single-leader control over Ethereum through techniques such as parallel processing of multiple blocks, synchronized release, and delayed execution. This ensures transactions can proceed freely while maintaining the fairness of the Ethereum network.

Braid Technical Features

The design principles of Braid primarily include three points: multi-proposer model, synchronized release, and delayed execution.

The Ethereum network enhances its censorship resistance in the multi-proposer model by incorporating multiple concurrent proposers. This approach significantly increases the cost of intervening in system transactions. Additionally, through the synchronized release mechanism, it ensures that all proposers make decisions based on the same information, thereby guaranteeing the system’s fairness and transparency.

Furthermore, the delayed execution feature introduced in the model allows multiple proposers to influence transactions before determining their final state, thus enhancing the stability and reliability of the entire Ethereum network. This design not only improves the system’s functionality but also provides network participants with a higher degree of trust and security assurance.

Advantages of Braid

At the aforementioned workshop, Max Resnick compared Braid to existing solutions such as LMD-Ghost and Mysticeti. LMD-Ghost, a fork choice rule for the PoS consensus algorithm CBC Casper, enables proposers to generate new blocks anytime. These new blocks are added to the chain with the highest weight, as determined by the fork selection rule (ensuring liveness). However, this approach necessitates the local maintenance of a highly branched decision tree to manage branch selection issues.

Mysticeti, on the other hand, is the consensus algorithm adopted by the Sui blockchain. It allows multiple validators to propose blocks in parallel, utilizing the full bandwidth of the network and providing censorship resistance. These characteristics of DAG-based consensus protocols require only three rounds of messages to commit blocks from the DAG, similar to pBFT, and reach the theoretical minimum. The commit rules allow voting and certification of block leaders, further reducing median and tail latency. The rules also tolerate unavailable leaders (when a leader node fails, the system automatically elects a new leader to take over its responsibilities) without significantly increasing commit latency.

LMD-Ghost (Source: Youtube）

Braid’s multi-proposer model allows Ethereum’s execution layer to collect block transactions generated by all sub-chains within a slot, forming an execution block. These transactions are then sequenced and executed according to predetermined rules, reducing the ability of a single entity to manipulate transaction records. One challenge this faces is the need for a deterministic sequencing rule.

Braid’s design does not introduce additional roles for incentives or penalties, but its “synchronized release” mechanism is difficult to implement, requiring coordination of multiple subchain synchronization and data processing.

Jonahb, a member of the Blockchain Capital team, pointed out a problem in the Braid mechanism: the “tip” mechanism has liquidity requirements, which affects user experience. Users will include two tip values (t, T) when submitting a transaction. If only one proposer includes a transaction, they will receive T; if multiple proposers include the transaction, they will split t.

Although users only need to pay the transaction fee, they need to have T available funds to make a credible commitment to the protocol that they can pay the fee T. Therefore, users need additional available liquidity T to conduct transactions. For example, if a user wants to sell $5 million worth of ETH due to concerns about upcoming interest rates and values censorship resistance at $1 million. This poses additional and vague liquidity requirements for participants, increasing the position’s value, thus hindering the user experience of on-chain finance.

To address this challenge, Jonahb suggested two potential solutions:

• Post-State Liquidity Verification: When submitting a transaction, users must provide evidence of sufficient funds to pay T after the transaction completes (e.g., having $1 million in liquidity post-transaction). This allows users to conduct transactions by proving future payment ability, even if they lack funds. However, this method faces challenges: proposers must predict the final transaction state, while most financial transactions involve shared states (e.g., multiple transactions using the same account balance). Consequently, proposers struggle to predict the post-transaction state before determining the transaction order. This necessitates designing specific proofs for different transaction types, limiting practical applicability.
• Anti-Censorship Assurance: Introduce third-party assurance providers (CI providers) to guarantee T for users. Users pay an assurance fee rT, where r is calculated based on the probability of transaction censorship. This alleviates the pressure on users to prepare large amounts of funds immediately and allows CI to remind users of high censorship risks if T is too low. However, building a market system between users and CI providers will take some time.

Regarding specific implementation, Braid employs a multi-block parallel operation model and a unified consensus protocol to ensure consistency and coordination among these blocks. This approach promotes decentralized transaction processing, effectively enhancing censorship resistance. Additionally, Braid has developed a “Finality Gadget” to handle transaction finality. This solution integrates and reorders all on-chain transaction sets, ensuring both finality and consistency of transactions.

Braid is still in its early stages as a proposal, and the market has yet to validate its technical implementation and user experience.

Braid (Source: Youtube)

Despite the flourishing of numerous blockchains today, Ethereum—the pioneer of programmable blockchains—continues to pursue decentralization, censorship resistance, and sovereign independence. As new technologies and reform strategies evolve, we remain optimistic and confident about blockchain technology’s future. By continuously exploring and balancing various solutions, we can ensure that technological progress meets current needs and adapts to future challenges.