一种高效的隐私集合交集协议

An Efficient Private Set Intersection Protocol

隐私集合交集计算已经被广泛地应用到各个行业中。随着数据量的不断增长,传统的隐私集合交集协议不再能有效地满足实际需求。主要从隐私保护的角度出发,设计满足实际需求的大数据隐私集合交集协议。结合布隆过滤器,构造了一个安全高效的隐私集合交集协议,该协议在半诚实敌手模型下安全通过实验测试。当集合大小为百万级时,协议在并行模式下的执行时间仅有15s,同时可达到128位的安全级别。

Private Set Intersection(PSI)has extensive practical applications.With the increasing of data volume,traditional PSI protocols can no longer meet the actual needs.From the point of view of privacy protection,an efficient private set intersection protocol over large-scale datasets is designed.We construct a secure and efficient PSI protocol over large-scale datasets in the server-aided setting based on Bloom Filter algorithm.Our protocol is secure against a semi-honest server.Furthermore,experimental results show that our protocol only needs around 15(128-bit security in parallel mode)over one million-element datasets.