半诚实模型下安全多方排序问题的研究  

Research on the Problem of Secure Multi-party Ranking Under Semi-honest Model

作　　者：肖倩[1] 罗守山[1] 陈萍[2] 吴波[3] 

机构地区：[1]北京邮电大学软件学院 [2]北京邮电大学电信工程学院,北京100876 [3]北京邮电大学理学院,北京100876

出　　处：《电子学报》2008年第4期709-714,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.60642008);西安电子科技大学综合业务网理论及关键技术国家重点实验室开放课题(No.ISN7-01)

摘　　要：安全多方排序问题是百万富翁问题的推广问题,用于n个参与方在不泄漏各方秘密输入的前提下比较出其输入在全体输入中按照一定顺序所处的位置.本文首先提出了半诚实模型下基于同态加密的安全两方排序协议.然后将该协议推广到多方排序的情况,并提出两种提高效率的改进算法.最后本文还提出了基于模糊贴近度的安全多方排序协议,并对这几个协议的安全性和效率做了分析、比较.Secure multi-party ranking is a problem generalized from the millionaires＇ problem, which can be used by n people to know about their secrets＇ order among all their inputs without leaking further information. Through the study on millionaires＇ protocol, we presented a secure two-party ranking protocol under semi-honest model based on homomorphic encryption. Then we generalized it to secure multi-party ranking,and we presented two algorithms whose efficiency are both improved.Finally, we gave a secure multi-party ranking protocol based on fuzzy nearness degree, and we analyzed the efficiency and security of these protocols.

关 键 词：百万富翁问题 安全多方排序 半诚实模型 同态加密 模糊贴近度 

分 类 号：TN918.1[电子电信—通信与信息系统]