信息论安全的3个基础外包计算协议及空间位置关系保密判定

Three basic information-theoretic secure protocols for outsourcing computing and privacy-preserving determination of spatial location-relation

现存的安全几何计算协议大多采用公钥加密方法保护数据隐私,计算成本较大。当计算能力不强的用户解决复杂问题时,效率往往较低。针对这些问题,避开公钥加密的方法,而是利用矩阵论中一些特殊函数的性质和随机数混淆的方法来保护数据隐私,并且为了进一步提高效率,将大量的用户计算外包出去。在此基础上,首先设计了常用的3个基础向量外包计算协议,分别是安全外包计算的向量模长计算协议,向量内积计算协议和向量夹角计算协议,并利用模拟范例证明了协议的安全性,然后利用这3个基础协议进一步解决了现实意义中如何保密判断空间面与面位置关系的问题,并给出了具体协议。最后,通过理论分析与实验仿真显示:由于协议没有使用公钥加密的方法,因此达到了信息论安全;并且由于外包计算的使用,为用户节省了更多的计算成本,取得了比较高的效率;此外,协议能够解决的问题也更加广泛,可作为新的云计算技术的基础协议应用到安全多方计算的其它分支中。

The existing protocols for secure geometric computation mainly make use of the public key encryption to protect the privacy of data,which is more costly.And there arises a problem of inefficiency for the users weak in computing power solve complicated problems.Aiming at these drawbacks,this paper combines the properties of some special functions in matrix theory with the random number to protect the data’s privacy instead of the public key encryption.In the meanwhile,numerous computation tasks are also outsourced in order to further improve the performance.On the basis of these techniques,we first design three basic protocols for vector calculation,including security outsourcing calculation of vector length,vector inner product,and vector angle with the security of our protocols proved with simulation paradigm.Then we employ them to privately determine the spatial location-relation of plane and the plane in the sense of real life.Finally,the analysis and comparison under theory and simulation experiments shows that our protocols achieve the information-theoretic security and a higher efficiency because of using the outsourcing computing rather than the public key encryption,as well as they can be applicable to solve more problems than the previous ones.The proposed protocols in this paper as the new techniques of cloud computing can be used in building-blocks in secure multi-party computation.