DPPR: A Dynamic, Privacy-Preserving Scheme for Reputation-Based Blockchain
Yun Hao, Jingyu Wang, Lixin Liu, Yongxing Du, Zetian Zhang, Hanqing Yang · 2024
Reputation-based blockchain systems have attracted widespread attention due to their ability to effectively combine key services such as energy trading and supply chain with blockchain technology. However, they face challenges in reliably evaluating the honesty and capabilities of consensus nodes, and are often vulnerable to slowly adaptive attacks against high-throughput nodes, which can undermine system reliability. To address these issues, we propose a dynamic privacy protection scheme, DPPR, for reputation-based blockchains. DPPR adopts a dynamic logistic regression model to assign reputation scores based on historical node behaviors, and controls the reputation score limit to enhance the reliability of node evaluation and maintain system balance. It also introduces confidential smart contracts leveraging fully homomorphic encryption for privacy-preserving reputation updates to prevent attackers from identifying high-reputation nodes. Simulation results show that DPPR can effectively deal with slowly adaptive attacks, reduce consensus delays in privacy-sensitive scenarios, and provide reliable node evaluation, showing comprehensive advantages.