How Far Can Decentralized Sequencers Go?

In Summary: Sequencers can address the scalability and performance limitations of the Ethereum network. They alleviate the burden on the Ethereum main chain by sorting and batch processing on-chain transactions, thus shifting most of the computational and data storage work to Layer 2. The decentralized sequencers’ capability ceiling depends on the underlying protocol and the network environment. While decentralized sequencers can improve the security and resilience of a system, there are still limitations and challenges.

Sequencers are a core component of rollup networks, responsible for crucial operations including receiving, sorting, and executing transactions, as well as submitting transaction data. If the only sequencer in a network fails or becomes unavailable, the whole network will stop processing transactions. However, many of the existing rollup solutions only have a single sequencer, making them far less decentralized than some centralized Layer1 alternatives. Therefore, the importance of decentralized sequencers is self-evident, and promising decentralized sequencers should effectively increase the decentralization of a network with optimized design and implementation.

Importance of Sequencers

Existing rollup solutions include zero-knowledge proofs (ZKPs) based zk-Rollups and optimistic execution-based Optimistic Rollups. These solutions are more scalable than monolithic Layer1s. But they still have their respective issues.

Issues with zk-Rollups:

Computational complexity: Using ZKPs to verify the correctness and legitimacy of transactions requires a significant amount of computational resources and time. This may result in transaction-processing delays and high computational costs.
Dependency on verifiability: Zk-Rollups rely on external verifiability, meaning that external supervisors are required to verify the correctness of ZK proofs. This may introduce trust issues and centralization risks.

Issues with Optimistic Rollups:

Reversibility: Optimistic Rollups work on an “optimistic” assumption that all transactions submitted are valid and conflict-free. However, if there are conflicts or invalid transactions, the entire system may have to roll back and re-execute, leading to uncertainties and processing delays.
MEV: Optimistic Rollups may face MEV (Maximal Extractable Value) issues, such as manipulated and unfair transaction ordering.

These issues reduce the performance and security of existing rollup solutions and may impact usability and user experience. The introduction of new designs like sequencers to improve the performance and decentralization of rollups is key to addressing these issues. Sequencers can increase throughput and compress transaction data. Specifically, they order transactions according to predetermined rules, which can not only enhance the efficiency and throughput of transaction processing but also reduce conflicts and competition among transactions. Sequencers also compress transactions, bundling multiple transactions into a single one, thereby reducing the scale of transaction data. Such compression helps reduce the costs for on-chain storage and transmission while improving the whole system’s efficiency.

Disadvantages of Centralized Sequencers

Most of the existing rollup service providers maintain their own centralized sequencers as it’s more convenient and cheaper. However, the disadvantages are also evident, including but not limited to susceptibility to censorship, excessive fees, and opportunities for capturing MEV maliciously.

Sequencer decentralization is considered an important direction for rollup maturing and a promising alternative to avoid the disadvantages of centralized sequencers. Decentralized sequencers are believed to offer more security, liveness, and censorship resistance. But they are not easy to design and implement. And the best solutions should make wise trade-offs among various factors, including performance, decentralization, and security.

Towards Decentralization

The current technology for decentralized sequencers is crude and can be improved by finding more effective ordering algorithms, implementing more robust validation mechanisms, creating smarter designs, etc. We’ve summarized below some meaningful decentralization approaches that are currently being explored. And as technology evolves, we expect decentralized sequencers to have higher throughput, faster confirmation speed, and lower latency as well as higher security and composability.

Proof-of-Authority (PoA): In this approach, a handful of entities are selected to take turns operating sequencers in a PoA system. The approach improves censorship resistance and has the lowest latency. But the drawback is it still faces the risk of single-point failure.
Based Sequencing: In this approach, there are no privileged sequencers, and anyone can submit batches to L2. Transaction ordering and block proposing are deferred to the Data Availability (DA) layer. The advantage is it inherits the liveness and censorship resistance of the DA layer. But the downside is the proceeds may leak to the base layer and it is more susceptible to MEV attacks. Also, the confirmation time is long.
Distributed Validator Technology (DVT): With DVT, the responsibilities of running a single sequencer can be distributed across a cluster of machines and node operators. Each node operator can sign independent attestations using their fractional share of the validator key. This approach offers the flexibility to be plugged into other solutions, but it introduces a little latency.
Shared Sequencing: This solution allows many rollups to share a single decentralized network of sequencers. This shared sequencer network processes transactions on several chains in parallel and provides cross-chain atomicity, real-time censorship resistance, and strong economic security at the sequencing layer. Shared sequencers have the network effect of serving multiple chains, but they are still limited by the L1’s data and transaction ordering throughput.
Bootstrapping New Sequencer Sets: This approach creates a decentralized sequencer group without permission using token incentive mechanisms. This approach involves creating a decentralized sequencer set permissionlessly by adopting token incentives. The advantage is increasing token utility, but the downsides are the latency and the difficulty of implementation for lesser-known rollups. Despite the benefits offered by decentralized sequencers, each technical approach has its own tradeoffs.

Potential Opportunities

Now, let’s take a step back and ask the question of whether decentralized sequencers can permanently solve Ethereum’s problems. Do they at the same time introduce hidden risks to Ethereum?

Firstly, for solutions that utilize the Ethereum L1 for sequencing such as based sequencing, the rollup system’s performance will be fundamentally limited by the throughput on L1 and can only alleviate computational bottlenecks and achieve small factor improvements in communication complexity.

Secondly, network conditions are also an important consideration. The stability and synchronicity of the network will directly impact the liveness and security of sequencers. Sequencers will lose liveness when the network behaves asynchronously and fail to respond to transactions timely. Only in networks with good synchronicity can sequencers maintain reliable liveness degrees.

Therefore, infrastructure could be a potential investment opportunity. That includes sequencer service providers, security auditors, cross-chain solution providers as well as governance and participation platforms, etc. The solutions offered by these infrastructure providers might solve the problems with existing decentralized sequencing approaches. But it should be noted that decentralized sequencing is only one possible direction that the Ethereum community is exploring to elevate performance and scalability. It’s not the only way. Also, new solutions will emerge as technology advances. We expect decentralized sequencing solutions to evolve in the following directions.

Multi-chain Interoperability: With the emergence of various blockchains and Layer 2 solutions, multi-chain interoperability could become an essential aspect of decentralized sequencers. Future sequencers may need to handle transactions across multiple chains simultaneously and achieve atomic composability to provide smoother user experiences and more powerful functionalities.
Stronger MEV Prevention and User Protection: Future sequencers would be able to reduce the impact of MEV and better protect users against monopoly pricing. This may include adopting random ordering mechanisms, reasonable transaction fee mechanisms, and improved privacy protection measures.
Enhanced Governance and Participation: To ensure the fairness and security of decentralized sequencers, future sequencers may introduce stronger governance and participation mechanisms. This can be achieved through token holder voting, validator elections, and decentralized decision-making by participants. More open and transparent governance mechanisms can promote community participation and drive the development of the system.

In conclusion, as decentralized sequencers continue to evolve, we expect to see more business model innovations. These may include different transaction fee models and sequencer-based data services and on-chain applications, and more. Innovative business models will provide more economic incentives for sequencers, thereby promoting their widespread adoption and sustainable development.