Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AQUAREUM: Non-Equivocating Censorship-Evident Centralized Ledger with EVM-Based Verifiable Execution using Trusted Computing and Blockchain

Ivan Homoliak, Mario Larangeira, Martin Peresini, Pawel Szalachowski

TL;DR

Aqua-reum tackles centralized-ledger limitations by fusing a trusted execution environment with a public blockchain to achieve verifiable, non-equivocating, censorship-evident operation. It leverages a Turing-complete VM inside the enclave and a history-tree-based ledger to enable efficient, incremental updates and logarithmic proofs, while publishing version transitions on-chain to deter equivocation. The work provides a formal UC-security argument and a practical prototype demonstrating throughput above $400$ TPS on commodity hardware, along with architectural considerations for censorship resolution and recovery from enclave failures. The approach offers a scalable alternative to fully decentralized ledgers for applications demanding strong verifiability, privacy-preservation during censorship-resolution, and lower client costs.

Abstract

Distributed ledger systems (i.e., blockchains) have received a lot of attention. They promise to enable mutually untrusted participants to execute transactions while providing the immutability of the data and censorship resistance. Although decentralized ledgers are a disruptive innovation, as of today, they suffer from scalability, privacy, or governance issues. Therefore, they are inapplicable for many important use cases, where interestingly, centralized ledger systems might gain adoption. Unfortunately, centralized ledgers have also drawbacks, e.g., a lack of efficient verifiability or a higher risk of censorship and equivocation. In this paper, we present AQUAREUM, a novel framework for centralized ledgers removing their main limitations. By a unique combination of a trusted execution environment (TEE) with a public blockchain, AQUAREUM provides publicly verifiable non-equivocating censorship-evident private and high-performance ledgers. AQUAREUM is integrated with a Turing-complete virtual machine (e.g., EVM), allowing arbitrary transaction processing logic, such as transfers or client-specified smart contracts. AQUAREUM is fully implemented and can process over 400 transactions per second on a commodity PC. Furthermore, we modeled AQUAREUM using the Universal Composability framework and proved its security.

AQUAREUM: Non-Equivocating Censorship-Evident Centralized Ledger with EVM-Based Verifiable Execution using Trusted Computing and Blockchain

TL;DR

Aqua-reum tackles centralized-ledger limitations by fusing a trusted execution environment with a public blockchain to achieve verifiable, non-equivocating, censorship-evident operation. It leverages a Turing-complete VM inside the enclave and a history-tree-based ledger to enable efficient, incremental updates and logarithmic proofs, while publishing version transitions on-chain to deter equivocation. The work provides a formal UC-security argument and a practical prototype demonstrating throughput above TPS on commodity hardware, along with architectural considerations for censorship resolution and recovery from enclave failures. The approach offers a scalable alternative to fully decentralized ledgers for applications demanding strong verifiability, privacy-preservation during censorship-resolution, and lower client costs.

Abstract

Distributed ledger systems (i.e., blockchains) have received a lot of attention. They promise to enable mutually untrusted participants to execute transactions while providing the immutability of the data and censorship resistance. Although decentralized ledgers are a disruptive innovation, as of today, they suffer from scalability, privacy, or governance issues. Therefore, they are inapplicable for many important use cases, where interestingly, centralized ledger systems might gain adoption. Unfortunately, centralized ledgers have also drawbacks, e.g., a lack of efficient verifiability or a higher risk of censorship and equivocation. In this paper, we present AQUAREUM, a novel framework for centralized ledgers removing their main limitations. By a unique combination of a trusted execution environment (TEE) with a public blockchain, AQUAREUM provides publicly verifiable non-equivocating censorship-evident private and high-performance ledgers. AQUAREUM is integrated with a Turing-complete virtual machine (e.g., EVM), allowing arbitrary transaction processing logic, such as transfers or client-specified smart contracts. AQUAREUM is fully implemented and can process over 400 transactions per second on a commodity PC. Furthermore, we modeled AQUAREUM using the Universal Composability framework and proved its security.

Paper Structure

This paper contains 63 sections, 7 theorems, 12 figures, 6 algorithms.

Table of Contents

  1. Introduction
  2. Proposed Approach
  3. Background
  4. Blockchain and Smart Contracts
  5. Trusted Execution Environment
  6. Merkle Tree
  7. History Tree
  8. Notation
  9. System Model and Overview
  10. System Model
  11. Desired Properties
  12. High-Level Overview
  13. Setup
  14. Operation
  15. Ledger Update Procedure
  16. ...and 48 more sections

Key Result

Theorem 1

Assuming a collision-resistant hash function, security of Merkle, incremental and membership proofs, and the Existential Unforgeability under Chosen Message Attack (EUF-CMA) scheme for $\mathbb{E}$, then $\Pi_{AQ}$ realizes $\mathcal{F}\xspace_{L}$ in the presence of $\mathcal{A}\xspace$ who corrupt

Figures (12)

  • Figure 1: Operation procedure of Aqua-reum ledger.
  • Figure 2: Aqua-reum components. Trusted ones are in green.
  • Figure 3: Protocol for setup ($\Pi_{S}$).
  • Figure 4: Protocol for normal operation ($\Pi_{N}$).
  • Figure 5: Protocol for receipt retrieval ($\Pi_{R}$).
  • ...and 7 more figures

Theorems & Definitions (13)

  • Theorem 1
  • Theorem 2
  • proof : Proof (Informal)
  • Theorem 3
  • proof : Proof (Informal)
  • Theorem 4
  • proof : Proof (Informal)
  • Theorem 5
  • proof : Proof (Informal)
  • Theorem 6
  • ...and 3 more