群智感知中基于区块链的共识节点招募方法

Blockchain-Based Consensus Node Recruitment Scheme in Crowdsensing System

随着移动终端的快速普及,群智感知系统得到了学术界的广泛关注.现有工作大多数依赖第三方平台来执行参与者选择和收集大规模数据,很难兼顾对参与者隐私的保护和数据质量评估的公平.针对该问题,本文提出了一种群智感知中基于区块链的共识节点招募方法.通过Pedersen承诺隐藏参与者的选举随机值,借助零知识证明系统和区块链技术保证了选举数据的隐私性和完整性,实现在参与者中公平且秘密地选举单个共识节点.考虑到单点故障的隐患问题,该方法支持定时重启选举,保持共识节点不断更新以抵御共谋攻击;还结合Pedersen承诺设计了更新阶段,安全定期更新选举随机值以抵御动态腐化.结合区块链对数据质量评估结果进行二次共识后,由共识节点基于贡献更新信誉和计算报酬,保证了参与者的奖惩公平并激励参与者提供高质量的任务数据,从而提高用户参与任务的积极性.仿真实验表明,该方法不仅实现了安全公平的选举过程,而且不会额外增加参与者的计算和存储负担.

With the rapid popularization of mobile terminals,crowdsensing systems have received extensive attention in academic circles.Nevertheless,most existing work relies on a trusted third-party to perform worker selection and collect data on an enormous scale,which is hard to balance protection of worker’s privacy and fair data quality assessment.To tackle these issues,a blockchain-based consensus node recruitment scheme in crowdsensing system(BNRS)is proposed.The BNRS realizes selecting a single consensus node in workers fairly and secretly through the Pedersen commitment to hide worker’s election random key and using the zero-knowledge proof system and blockchain to ensure the privacy and completeness of data in elections.Considering the hidden trouble of single point failure,the BNRS supports restarting election regularly to maintain consensus node updates to resist collusion attacks.The BNRS also designs an updating phase based on the Pedersen commitment,which updates election random key regularly and safely to resist dynamic corruption.After reaching twice consensus on data quality evaluation results,consensus node updates reputation and calculates remuneration based on contributions to ensure fair rewards and punishment for workers and encouraging them to provide high-quality data,thereby improving the enthusiasm of workers to participate in task.Simulation results indicate that the BNRS not only achieves a secure and fair election process,but also does not increase additional computational and storage burden on workers.