Parallel Byzantine Fault Tolerant (BFT) protocols based on committee-based sharding improve scalability but weaken safety since smaller node groups are responsible for consensus. Recent approaches integrate trusted execution environments (TEEs) into parallel BFT frameworks to enhance safety. While the scalability and safety issues are addressed by trusted parallel BFT, existing committee configuration methods often rely on randomized assignment, which can degrade performance. This paper proposes a committee configuration optimization (CCO) model based on mixed integer programming to improve transaction performance for trusted parallel BFT. The model considers communication delays and node failure rates to determine an optimal committee configuration that minimizes transaction latency under both normal operations and scenarios of trusted hardware failures. We integrate CCO into a trusted parallel BFT protocol and evaluate the performance on Microsoft virtual machines. Experimental results demonstrate 15% and 21% improved transaction throughput under normal operations and fallback process, respectively, highlighting the benefits of optimization-driven committee configuration in trusted parallel BFT systems.

翻译：基于委员会分片的并行拜占庭容错（BFT）协议通过分片提高了可扩展性，但由于共识由规模更小的节点组负责，安全性有所削弱。近期研究将可信执行环境（TEEs）集成到并行BFT框架中以增强安全性。尽管可信并行BFT解决了可扩展性与安全性问题，但现有的委员会配置方法通常依赖随机分配，这可能降低系统性能。本文提出一种基于混合整数规划的委员会配置优化（CCO）模型，以提升可信并行BFT的交易处理性能。该模型综合考虑通信延迟与节点故障率，以确定最优的委员会配置，从而在正常运行和可信硬件故障场景下最小化交易延迟。我们将CCO集成到一个可信并行BFT协议中，并在微软虚拟机上进行了性能评估。实验结果表明，在正常运行和回退处理过程中，交易吞吐量分别提升了15%和21%，突显了优化驱动的委员会配置在可信并行BFT系统中的优势。