Parallel Byzantine fault tolerance consensus based on trusted execution environments

Ensuring consistency and reliability in distributed systems is crucial for their adoption. Blockchain technology offers a robust framework for these systems, characterized by decentralization, immutability, auditability, and traceability. In public blockchains, consensus algorithms like Proof-of-Work (PoW) are foundational for securing transactions but are criticized for high energy consumption and limited transaction throughput, making them less suitable for high-frequency environments. In contrast, consortium blockchains rely on Byzantine Fault Tolerance (BFT) algorithms to address these issues. However, existing Practical Byzantine Fault Tolerance (PBFT) protocols still face challenges in performance, scalability, and fault tolerance. This paper introduces a novel Parallel Byzantine Fault Tolerance protocol, TEP-BFT, leveraging Trusted Execution Environments. The TEP-BFT protocol utilizes a Unique Sequential Identifier Generator (USIG) based on Intel Software Guard Extensions (Intel SGX) to generate unique identifiers, thus ensuring the monotonicity, uniqueness, and order of messages. This innovation reduces the requisite number of communication phases and replicas, substantially enhancing the efficiency and fault tolerance of the consensus process. Moreover, the protocol implements parallel processing strategies both inter-thread and intra-thread to augment the throughput of the blockchain system significantly. Our experimental and performance analysis indicates that TEP-BFT achieves an optimal balance among performance, scalability, and fault tolerance, surpassing other BFT protocol variants. This advancement positions TEP-BFT as a superior choice for blockchain systems in scenarios requiring rapid and frequent transaction processing, marking a significant step forward in the evolution of blockchain consensus mechanisms. Our code is made public available at: https://github.com/SICC-Group/TEP-BFT.git.