无茫然第三方的安全两方向量优势统计协议

安全两方向量优势统计是一类特殊的安全多方计算问题,用于统计两方在不泄露各自私有向量信息的前提下,满足大于关系的分量数目。但现有的安全两方向量优势统计协议都依赖于茫然第三方,协议的安全性和效率较低。为此,在半诚实模型下,利用同态加密算法和向量叉积协议,提出一个无需茫然第三方支持的两方向量优势统计协议。理论分析结果表明,该协议无需茫然第三方即可提高协议的安全性。该协议的通信轮数为2,通信代价较低。在此基础上,将该协议应用于安全两方向量分量和的排序,也能显著提高排序性能。

Secure Two-party Vector Dominance Statistic Protocol Without Oblivious Third Party

Secure two-party vector dominance statistic is a special secure multi-party computation problem, which can be used by two parties to get the number of ai〉biwithout leaking their private vector information. The protocol security and efficiency is very lower in existing secure two-party vector dominance statistic protocol rely on oblivious third party. To address this situation, this paper proposes a new secure two-party vector dominance statistic protocol without oblivious third party based on homomorphic encryption and vector cross protocol in semi-honest model. Theoretical analysis shows that this protocol doesn＇t include oblivious third party so that its security is improved, and only needs two rounds communication that leads to the reduction of communication complexity. The performance of ranking protocol is improved significantly when the new secure two-party vector dominance statistic protocol is applied in secure components sum of two vectors ranking.