Sequencer is a key technology in the field of cryptocurrency, which is used to sort transactions and create blocks. Before the block is confirmed, the pre-confirmation information will be sent to the user.

first level title

What is a "sequencer"?

Sorters allow L2 to run efficiently by aggregating many L2 user transactions off-chain and submitting them as an aggregated single transaction to the main chain L1. This way, the cost of this commitment can be amortized across all user transactions in the set. Sorters can also compact collections to further save on main chain data availability costs. Overall, it is an essential component of L2.

However, the sorter has control over the ordering of transactions in the collection. The sorter can choose not to include user transactions, and the sorter can also extract MEV (Maximum Yield Realizable) in the collection through standard reordering and insertion extraction methods. They actually have preferential write access to the expansion. Notably, since the orderer can interact with the contract, only error-free transactions can be reliably enforced through on-chain mechanisms. And there are bugs that may fail when forcing the sort.

This makes the sequencer a semi-trusted role for scaling users. Orderers can delay user access and extract value from user transactions. Further restricting the behavior of the orderer through decentralization is a topic of active research.

first level title

Sequencer centralization risk

Currently, the sorter in optimistic scaling still has some problems in terms of distribution. Since the orderer usually plays a centralizing role, there are the following centralization risks:

• Weak censorship resistance: Unlike the near-infinite number of distributed nodes on the main chain, centralized orderers may not be able to ensure that user transactions will be included on-chain. A centralized coordinator controlled by a legal entity may selectively exclude specific transactions due to regulatory requirements. Although there are other mechanisms to solve the weak censorship problem of optimistic scaling (such as forced exit, escape channel, include list or adding threshold encryption, etc.), we still need to accept the assumption that the centralized orderer is likely to have weak Anti-censorship capability.

• Weak Liveness: A centralized orderer may not be designed to handle the computational processing and proof generation required to keep the system running all the time. RPC or orderer downtime due to hardware failure or mass spam from validators or bots (e.g. Arbitrum Token Launch, Optimism Delay) can lead to less active scaling.

• MEV benefits: Current centralized orderers typically follow a first-come, first-served transaction ordering rule. Additional trust is required to ensure they are not extracting MEV from user transactions through node privileges, or that the third-party ordering services they employ (such as Chainlink FSS) do not behave maliciously.

first level title

Sequencer layout

Vitalik Buterin proposes several ways to build a decentralized sorter. These include sorter/block auctions, PoS-based random selection, and DPoS voting, among others. However, most solutions focus on determining which participants have the right to propose the next block or sequence of blocks, often ignoring the ordering mechanism itself.

The goal of PBS is to protect proposers from centralized maximized transaction value (MEV), facilitate block builder competition, enhance bidder privacy, and eliminate negative externalities. However, unlike the first-layer solutions (L1), L2's PBS faces challenges such as privacy, latency, and cross-chain MEV. One way to address the privacy issue is to use Flashbots' SUAVE protocol, and combining SUAVE with a shared orderer is a potential PBS solution for L2.

For Aztec's PBS --"Prover-Builder-Separation (verifier builder separation)"instead of"proposer-builder separation". Proposers of Aztec build blocks with pending transactions from the mempool, blocks include ordering commitments, rewards to provers, and the amount Aztec burns. It should be noted that the proposer of Aztec actually plays the role of builder and proposer.

Aztec's PBS separates the power of transaction ordering (builders) and block inclusion (validators), a separation that prevents monopoly of block generation. Proposers then collect votes and create a block record indicating the distribution of validation tasks among multiple validators for a particular block. This is important to keep the validation task decentralized, as validator participation becomes an indicator of winning blocks.

Additionally, they use the TARGET_PROVERS count to increase the cost for the attacker to maintain the manipulation mechanism. One problem with this model, however, is that an attacker can avoid being penalized if they allow validators to be included and generate proofs for only a small fraction of blocks, leaving the majority to a single validator.

Throughout the process, multiple proposed blocks will be ranked through a voting phase, and the block with the highest number of votes will become the head of the chain. However, this model can lead to"griefing"Attack, where validators vote for blocks but do not generate proofs. Aztec suppresses this by introducing Slash and Redundancy mechanisms. Additionally, SUAVE can provide privacy protection and potentially decentralized block ordering as an Aztec builder.

There are a few other projects building shared sorters, including:

Espresso, plans to leverage EIgenLayer's ETH re-staking as a security model;

Astria, whose sequencer differs from Espresso in that it does not execute transactions, has PBS built in, and builds a Rollup on top of Celestia and Rollkit;

Radius, whose orderers focus on reducing harmful MEV through encrypted transactions, maintains a set of orderers and randomly selects one in each epoch.

Summarize

Summarize

With the continuous development and innovation of blockchain technology, the working mechanism of the decentralized sorter will continue to evolve and improve. This will provide users with a more secure, reliable and efficient trading experience while protecting them from manipulation and unfair practices by centralized institutions.

In the future, we can expect to see more innovative solutions and projects emerge to solve the challenges in the field of sequencers. With the advancement of technology, issues such as privacy protection, transaction speed, and cross-chain compatibility will be better resolved.

The development of a shared sequencer will enable different Rollups to work together and provide composability and flexibility to meet the needs of different industries, applications, and use cases. As shared sorters continue to mature and spread, we can foresee the emergence of thousands of decentralized sovereign Rollups, providing users with more choices and better services.

In conclusion, through continuous research and innovation, we have reason to believe that the future decentralized orderer will be a key component in building a safe, efficient and fair blockchain ecosystem. They will promote the further popularization and application of blockchain technology, bringing a more open and inclusive financial and digital experience to users around the world.