Exploring the Principles, Current Status, and Future Development of Layer 2 Sorters

The main source of income for Layer 2 is the Gas fees paid by users. After deducting the costs of submitting data to Layer 1, the remaining amount is basically pure profit. According to data statistics, the profit of a certain Layer 2 project in the second half of 2023 was approximately 5.23 million USD, another project's annual profit reached 16.5 million USD, and a third project's profit from March to December soared to 22.24 million USD.

Behind these huge profits is a close relationship with the sequencers that operate Layer 2 projects. So, what is a sequencer? How does it work in Layer 2? What problems do centralized sequencers currently face? How will sequencers evolve in the future? This article will delve into these questions.

Principle of the Sorter

The sorter plays a key role in Layer 2, mainly functioning to receive user transactions, execute transactions, and then submit the sorted and compressed batches of transactions to Layer 1.

The sorter can be likened to a bus driver. In the past, users traded directly on Ethereum, like driving into the city themselves, which often led to congestion. Layer 2 is like the public transportation system, where the sorter is the bus driver. The driver tells everyone that they don't need to drive themselves; they just need to pay a small service fee to reach their destination, saving both money and effort. To maximize the use of space inside the bus, the driver will try to fill it with passengers as much as possible before departing, and will also rearrange the passengers to ensure they are seated closely together.

Who can run the sorter?

Common solutions include:

1. Centralized Sorter: Exclusively operated by the Layer2 team or a designated organization, it is highly efficient and cost-effective, making it the team's preferred solution.

2. Permissionless Sorter: Anyone can sort and submit to Layer 1. It seems fair, but the drawbacks are obvious. The sorter is different from Layer 1 miners or validators and does not enhance the final security. When multiple parties submit batches simultaneously, only one will be included, resulting in wasted resources.

What is the sorting criteria?

There are usually two ways:

1. First come, first served: Similar to taking a seat on the bus, transactions that are submitted first will be prioritized.

2. Sort by Gas fees: Transactions that need to be processed urgently can pay more fees to be prioritized for packaging.

Mainstream Layer 2s mostly adopt the first method. In fact, there are no strict regulations on the sorting method for Layer 2; the sorters can theoretically decide on their own sorting method.

Can the sorter act maliciously? How to prevent it?

In theory, a sorter can act maliciously by falsely reporting success after canceling a transaction or by embedding malicious transactions to transfer users' assets.

To prevent malicious actions, different Layer 2 solutions have implemented various constraints.

• Optimistic Rollup uses fraud proofs, and if no one proves the data is incorrect during the dispute period (usually one week), the submitted data will become permanently unchangeable.

• ZK Rollup uses validity proofs to validate batches in real time, and upon passing, it receives final confirmation on Layer 1 without a dispute period.

Problems Caused by Centralized Sorters

Currently, mainstream Layer 2 solutions often adopt centralized sequencer schemes operated by official or affiliated organizations. Although this facilitates management, improves efficiency, and generates revenue, it still raises concerns among users:

1. Weak censorship resistance: Operated by a single entity, unable to compare with the large number of validators in Layer 1. May be required to remove transactions or set up blacklists due to regulatory requirements.

2. Low activity: High risk of single point of failure. A large number of transaction requests may lead to system downtime.

3. Obtaining improper MEV profits: Sorters may manipulate the order of transactions to gain additional profits. Practices such as private mempools have also raised concerns.

Future Development

Mainstream Layer 2 solutions have recognized the issues of centralized sequencers and have proposed decentralized solutions, but currently, they are only at the documentation or white paper stage. Compared to decentralizing power and profits, they are more focused on improving network performance and ecological development.

Decentralized Sorter Solution

1. Geographic Decentralization: Deploying multiple validators in different locations around the world, operated in turn by reputable organizations.

2. Sorter Auction: Auctions conducted through smart contracts, with the winning party required to stake a deposit.

3. Leader Election: Staked tokens participate in a random draw, with the probability proportional to the amount staked.

4. Based Rollup: Layer 2 transaction sorting led by Ethereum validators, with high technical difficulty and still unresolved issues.

Shared Sorter

Cancel the dedicated Layer2 exclusive sequencer for a single Layer2, allowing multiple Layer2s to share a third-party sequencer network. Advantages include achieving cross-Layer2 atomic composability and preventing MEV extraction, among others. Currently, multiple projects are establishing shared sequencer networks.

Summary of Thoughts

Whether decentralized orderers or shared orderers can perfectly solve the centralization problem remains to be seen. The MEV issue has become quite serious in permissionless networks like Ethereum; will decentralized Layer 2 orderers fall into the same trap? While shared orderers can enhance interoperability, if they are widely used, will they trigger new centralization issues?

The decentralization of blockchain is a long process. The sorter, as a key component of Rollup, has attracted much attention, and it is believed that through continuous exploration, the current issues will eventually be properly resolved.