欢迎访问南京邮电大学计算机学院、软件学院、网络空间安全学院!   |  联系我们
首页 > 学术交流 > 文章详情

2019年6月3日学术报告——万志国

来源: 计算机学院 | 发表时间: 2019-06-02 | 浏览次数: 13

报告题目:Efficient Decentralized Privacy-Preserving Usage-based Insurance for Vehicles

报告人:万志国

时间:  201963日(周一)10:00--11:00

地点:仙林校区计算机学科楼422会议室

主办单位: 计算机学院、软件学院、网络空间安全学院

摘要:

Compared with traditional insurance schemes, Usage-based Insurance (UBI) for vehicles is more economic and accurate for drivers since it calculates insurance premiums depends on how vehicles are driven. UBI also can incentivize drivers to drive safely since it reduces insurance premiums for safe drivers. However, UBI requires sensitive driving data to determine insurance premiums, and this could result in serious privacy breach for drivers. Meanwhile, existing UBI solutions rely on a centralized entity (the insurance company) to manage insurances. In this paper, we design a decentralized and privacy-preserving UBI scheme, called DUBI, based on the blockchain technology. In our scheme, a smart contract (Decentralized Application, DApp) running over the blockchain serves as a ``decentralized insurance company, while the driver continuously uploads his/her committed driving data to the blockchain. Periodically, the driver submits an accumulative driving result with a zero knowledge proof to the smart contract, which verifies the proof and calculates the insurance premium from the submitted result. In this way, DUBI achieves both security and privacy without relying on any centralized entity or any trusted tamper-proof hardware. We provide security analysis and performance evaluation for DUBI based on an implementation using Ethereum. The results show that DUBI is highly efficient in processing UBI insurances in both storage and computation.


万志国,山东大学计算机学院副教授,IEEEACM及中国计算机学会会员,中国计算机学会区块链专委首任委员,主要研究方向为区块链、云计算、大数据、物联网安全和隐私保护。20002002年分别获清华大学汽车工程、企业管理、软件工程学士学位,2006年获新加坡国立大学计算机学院博士学位,2006-2008年在比利时鲁汶大学电子工程系从事博士后研究工作。2008-2015年在清华大学软件学院工作,作为项目负责人负责国家自然科学基金项目、教育部留学回国人员基金项目、清华大学-鲁汶大学双边合作项目等多项项目。已发表学术文章40多篇,包括INFOCOMIEEE TDSC, IEEE TIFS等顶级国际会议和国际期刊。Google Scholar引用次数达1900余次,单篇引用次数达400+次。