EPF Final Updates

Topic:
Research project on decentralized prover mechanisms of zk-rollups
Ethereum Protocol Fellowship (July 2023 - Present)

Team: Me, Nilu and Norbert

Mentor:
Barnabé Monnot, Research Scientist at the Robust Incentives Group at the Ethereum Foundation

Project Overview:

The Ethereum Protocol Fellowship (EPF) research project, conducted from July to November 2023, centered on decentralized prover mechanisms for zk-rollups. Led by mentor Barnabé Monnot, our team—comprising Nilu, Norbert, and myself—explored the intricacies of zk-rollups, aiming to enhance Ethereum's scaling strategy. The project involved a comprehensive review of prover models, incentive mechanisms, and zk-L2 architectures. Our research culminated in the formulation of design criteria, followed by practical implementation through simulation and code. Notably, our prover model was submitted in response to Aztec's RFP, aligning our findings with real-world needs. The ongoing simulation analysis and integration with Aztec's proposal underscore the project's impact on Ethereum's scalability and decentralization journey.

Analysis and Review:
Our thorough exploration of decentralized prover mechanisms unveils a paradigm shift in the zk-rollup landscape. Key industry players like Aztec, Starknet, Taiko, and Scroll are steering towards decentralization, aligning with the evolving trends and challenges.

This ongoing research aims to craft an optimal prover network, fostering liveness and a competitive market. Our evaluative framework, rooted in core criteria like decentralization and cost, guides this quest, ensuring a delicate balance between competitiveness and decentralization.

Resources created by us:

• awesome-prover-mechanisms
• awesome-rollup-economics

Design Specification:
Our approach to prover mechanism design involves multi-objective optimization, seeking a set of design parameters that optimize key objectives while adhering to specified constraints. The following objectives and constraints frame our optimization problems:

1. Cost Optimization:

• Minimize Computational Costs for Provers
• Ensure Prover Operational Costs ≤ Prover Earnings

2. Liveness and Scalability:
3. Security
4. Honest behaviour
5. Censorship resistance
6. Permissionless entry
7. Transparency and fairness

Each criterion is accompanied by specific metrics and mathematical formulations, providing a comprehensive framework for evaluating and optimizing our prover mechanism. This structured approach ensures a robust, secure, and efficient decentralized prover network, aligning with our overarching goals and priorities.

Refining Prover Design:
As our project advanced, we shifted our focus towards refining the optimal prover design. Engaging in extensive discussions with our mentor Barnabé and collaborating closely with the team, we identified key priorities for our prover strategy, with a specific emphasis on decentralization, cost-effectiveness, liveness, and incentivizing honest behavior.

Following in-depth discussions and iterative refinements, our prover mechanism underwent its final evolution. Streamlining certain elements, introducing novel features, and aligning with our core priorities resulted in a comprehensive and pragmatic prover strategy. This ultimate design seamlessly incorporates staking with slashing to drive financial interest, employs prover reputation scores to enhance reliability, and integrates a random selection mechanism. To address potential prover failures, an emergency protocol is implemented to minimize reorganizations. Additional features, including proof batching, distributed proving, and the integration of liquid proving pools, further fortify the robustness of our prover mechanism. This culminating effort marks a significant milestone in our research journey.

Simulation Modeling for Prover Networks:
In our research, we constructed an agent-based simulation model using Python and the cadCAD framework. This model captures the dynamics of how provers interact, handle transactions, and respond to diverse network scenarios. The simulation includes aspects like malicious behavior, random system failures, and random selection of provers based on reputation. This approach facilitates a comprehensive analysis, enabling us to optimize the system, identify vulnerabilities, and enhance robustness and efficiency. The model processes random transactions, calculating user value and cost based on data size and unprocessed transactions.

Github Repo

Aztec's call for proposals seeking decentralized prover coordination solutions: Here

Proposal submitted to the Aztec Team:
Decentralized Prover Network (Staking, Reputations and Proof Races)

We conducted a thorough examination of Fernet, the chosen sequencer model by the Aztec team, along with a comprehensive review of other prover coordination proposals(here) already submitted to the Request for Proposal (RFP).

Future Endeavors:

As the Ethereum Protocol Fellowship (EPF) research project on decentralized prover mechanisms for zk-rollups approaches its conclusion, there are crucial outstanding items and areas of future work that demand attention and exploration. The forthcoming steps are integral to enhancing the comprehensiveness and practicality of the research findings.

1. Completion of Simulation:
   The simulation aspect of the project is still in its early stages, and there remains a need to diligently analyze how the devised prover mechanism behaves under diverse scenarios. This includes examining its performance under varying transaction loads, assessing its responsiveness in both normal and emergency modes, and evaluating its adaptability with different numbers of active provers in the network.

2. Refinement and Optimization:
   Iterative refinement of the prover mechanism is essential based on insights gained from the ongoing simulation. Optimization efforts should focus on fine-tuning parameters, enhancing efficiency, and addressing any unforeseen challenges or bottlenecks that may emerge during simulation experiments.

3. Integration with Aztec's Proposal:
   The submitted proposal to Aztec for decentralized prover coordination presents an exciting opportunity. Continued collaboration with the Aztec team is vital, and adjustments or enhancements to the prover model may be required based on the feedback and insights received during the proposal evaluation process.

4. Thorough Documentation:
   Comprehensive documentation of the finalized prover mechanism, including its design rationale, parameters, and simulation results, is crucial. This documentation serves not only as a record of the research but also as a valuable resource for the broader Ethereum community and researchers interested in decentralized prover models.

5. Continued Engagement and Impact Assessment:
   Ongoing engagement with stakeholders, including the Ethereum community and research teams, is essential for refining the decentralized prover mechanism. This involves addressing feedback, fostering discussions, and staying updated on the dynamic landscape of decentralized sequencers. Simultaneously, assessing the long-term impact of the prover mechanism on Ethereum's ecosystem and creating educational resources ensures its relevance, adoption, and contribution to the broader Ethereum scaling solutions.

In essence, the journey doesn't conclude with the research project's formal completion; rather, it extends into the practical implementation, community integration, and continuous refinement of the decentralized prover mechanism within the broader Ethereum landscape.

Conclusion:

In concluding our Ethereum Protocol Fellowship (EPF) research on decentralized prover mechanisms for zk-rollups, our journey has been both enlightening and impactful. Collaborating with my team—Nilu and Norbert—under the guidance of Barnabé Monnot, we delved into the intricacies of zk-rollups, aiming to enhance Ethereum's scaling strategy.

Our comprehensive analysis covered prover models, incentive mechanisms, and the architectures of various zk-L2s. The distilled findings led to the formulation of crucial design criteria, including decentralization, liveness, censorship resistance, and security.

Taking a practical approach, we ventured into simulation and code implementation. The ongoing analysis aims to provide real-world insights into our prover mechanism's behavior, ensuring its adaptability and robustness.

A pivotal moment arose when we aligned our research with Aztec's Request for Proposal (RFP), submitting our prover model. Looking ahead, refining the simulation, integrating with Aztec's proposal, and ongoing optimization underscore the continuous evolution of our research.

In essence, this EPF project marks not just a culmination but a beginning. The impact of our contributions extends beyond this research, influencing Ethereum's trajectory in scalability, security, and decentralization. As we conclude this chapter, the dynamic nature of the field ensures that our insights resonate in the ongoing narrative of Ethereum's growth.