【摘要】 传统空间数据共享计算需将数据集成后再分析,而数据提供者往往不想公开数据而拒绝分享。安全多方计算(MPC)用于解决互不信任的多方之间数据协同计算问题,能够为解决数据孤岛问题提供方案。本文以交叉K函数为例,探索空间数据安全共享方法,在原始数据不被泄露的前提下,实现目标函数的空间运算。搭建的原系统角色包括前端服务、控制服务、MPC代理服务及MPC发起方和MPC接收方。计算过程中,在双方认可的信任机制下,MPC发起方和接收方的数据明文始终保留在本地,双方均无法获知对方任何原始数据。最后对比了此方法和传统计算平台的时间效率和准确度。试验表明,本文方法可为空间数据安全共享提供一种全新途径,并为进一步的效率改进研究打下了基础。更多还原

【Abstract】 There is a surge in spatial data sharing computation research all over the world. The traditional method,however,should transmit some original data during the calculation,which may lead to data leakage,which the data providers are concerned. An effective way to eliminate data isolation is secure multi-party computation( MPC) framework that can solve the calculating problems between two unreliable parties. In this paper,we describe a new way of secure sharing of spatial data,based on the MPC framework,to implement cross K function,a typical point pattern analysis,which has been set in a prototype. From the front-end services and control services to MPC-proxy and MPC-application( including MPC-initiator and MPC-recipient),all roles in the prototype can get the result without transferring any original data. Finally,we compare the efficiency and accuracy of proposed method with the traditional computation platform. Experiments show that this method can provide a new approach for the safe sharing of spatial data and lay a foundation for further research on efficiency improvement.更多还原