Research on improved HotStuff consensus algorithm utilizing SM2 ring signature and reputation
The HotStuff primary node replacement algorithm exhibits regularity, enabling Byzantine nodes to pre-lock the primary node of the subsequent round, facilitating adaptive attacks that degrade the performance of the next consensus cycle. The proposed solution involves ensuring that the primary node selection process is randomized, thereby diminishing the likelihood of Byzantine nodes being chosen as the primary node. This paper presents a dynamic nodereputation system that progressively diminishes the reputation value of Byzantine nodes in each consensus round; it optimizes HotStuff by introducing Improved-HotStuff, which substitutes threshold signatures with SM2 linkable ring signatures. Additionally, it proposes a primary node replacement algorithm that integrates random election with reputation, preventing repetitive and false voting by Byzantine nodes. A novel <i>PreNewView</i> phase has been added to facilitate random primary node replacement, effectively addressing the adaptive attack issue. Security analysis and experimental outcomes demonstrate that Improved-HotStuff maintains a communication complexity of O(<i>n</i>), significantly reduces the success probability of adaptive attacks, and ensures both security and liveness. Under attack conditions, Improved-HotStuff can enhance throughput and reduce delay by over 20% compared to the original HotStuff algorithm.
Authors
- Qinan Li (ORCID: https://orcid.org/0000-0001-7818-4565)
- Kaiwen Zou
Institutions
- Lanzhou Jiaotong University (CN)
Publication Details
- Published
- 2025-04-25
- DOI
- https://doi.org/10.1117/12.3060737
- Primary Topic
- Advanced Computing and Algorithms
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Natural Science Foundation of Gansu Province