TBFT: Understandable and Efficient Byzantine Fault Tolerance using Trusted Execution Environment
share
While the requirements for reliability increasing rapidly in distributed systems, byzantine fault tolerance protocols have attracted more and more attention for the ability to tolerate arbitrary failures. Although traditional BFT protocols have achieved significant performance improvements after decades of research, they still cannot meet the increasing scalability requirements. Thanks to the recent development of trusted execution environment, researchers have opportunities to improve BFT protocols with hardware assistance, including increasing the ratio of byzantine nodes that the system can tolerate and improving the scalability of the system. However, existing TEE-based BFT protocols are mostly designed via improving traditional BFT protocols and using complex operations to overcome security issues introduced by TEE, which makes the protocols difficult to understand and implement. In fact, after introducing TEE, many byzantine failures can be eliminated naturally so that the adversary assumption of TEE-based BFT is more similar to CFT rather than traditional BFT. Therefore, it would be better to design a TEE-based BFT protocol on the basis of CFT protocols to inherit the advantages of CFT, i.e., understandable and efficient. In this paper, we summarize the key differences between TEE-based BFT and CFT protocols and propose four principles to help bridge the gap between TEE-based BFT and CFT. Based on these principles, we present TBFT, a novel TEE-based BFT protocol with inspirations from CFT protocols, which is simple and understandable enough to be applied in practice. Besides, we make comprehensive improvements to TBFT for both performance and security, including pipeline mechanisms, TEE-assisted secret sharing, and trusted leader election. Our evaluation shows that TBFT has better performance and scalability compared to previous protocols.
READ FULL TEXT