Fully homomorphic encryption is faced with two problems now. One is candidate fully homomorphic encryption schemes are few. Another is that the efficiency of fully homomorphic encryption is a big question. In this pap...Fully homomorphic encryption is faced with two problems now. One is candidate fully homomorphic encryption schemes are few. Another is that the efficiency of fully homomorphic encryption is a big question. In this paper, we propose a fully homomorphic encryption scheme based on LWE, which has better key size. Our main contributions are： （1） According to the binary-LWE recently, we choose secret key from binary set and modify the basic encryption scheme proposed in Linder and Peikert in 2010. We propose a fully homomorphic encryption scheme based on the new basic encryption scheme. We analyze the correctness and give the proof of the security of our scheme. The public key, evaluation keys and tensored ciphertext have better size in our scheme. （2） Estimating parameters for fully homomorphic encryption scheme is an important work. We estimate the concert parameters for our scheme. We compare these parameters between our scheme and Bral2 scheme. Our scheme have public key and private key that smaller by a factor of about logq than in Bral2 scheme. Tensored ciphertext in our scheme is smaller by a factor of about log2q than in Bral2 scheme. Key switching matrix in our scheme is smaller by a factor of about log3q than in Bra12 scheme.展开更多
分析以往格上基于身份的全同态加密方案,指出方案效率低的根本原因在于陷门生成和原像采样过程的复杂度过高,为此提出一种新的解决方案。先将新型陷门函数与对偶容错学习(LWE,learning with errors)算法有机结合,构造一种新的格上基于...分析以往格上基于身份的全同态加密方案,指出方案效率低的根本原因在于陷门生成和原像采样过程的复杂度过高,为此提出一种新的解决方案。先将新型陷门函数与对偶容错学习(LWE,learning with errors)算法有机结合,构造一种新的格上基于身份的加密方案;再利用特征向量方法转化为格上基于身份的全同态加密方案。对比分析表明,所提方案的陷门生成复杂度显著降低,原像采样复杂度约降低为以往方案的1/3,SIVP近似因子缩小为以往方案的1/m^(1/2)。在标准模型下,方案安全性归约至判定性LWE的难解性,并包含严格的安全性证明。展开更多
本文提出一种基于公钥密码体制(Number Theory Research Unit,NTRU)选择明文攻击(Chosen Plaintext Attack,CPA)可证明安全的全同态加密方案.首先,对NTRU的密钥生成算法进行改进,通过格上的高斯抽象算法生成密钥对,避免了有效的格攻击,...本文提出一种基于公钥密码体制(Number Theory Research Unit,NTRU)选择明文攻击(Chosen Plaintext Attack,CPA)可证明安全的全同态加密方案.首先,对NTRU的密钥生成算法进行改进,通过格上的高斯抽象算法生成密钥对,避免了有效的格攻击,同时,没有改变密钥的分布.然后,基于改进的NTRU加密算法,利用Flattening技术,构造了一个全同态加密体制,并在标准模型下证明方案是选择明文攻击不可区分性IND-CPA安全的.展开更多
基金The first author would like to thank for the Fund of Jiangsu Innovation Program for Graduate Education，the Fundamental Research Funds for the Central Universities,and Ningbo Natural Science Foundation，the Chinese National Scholarship fund,and also appreciate the benefit to this work from projects in science and technique of Ningbo municipal.The third author would like to thank for Ningbo Natural Science Foundation
文摘Fully homomorphic encryption is faced with two problems now. One is candidate fully homomorphic encryption schemes are few. Another is that the efficiency of fully homomorphic encryption is a big question. In this paper, we propose a fully homomorphic encryption scheme based on LWE, which has better key size. Our main contributions are： （1） According to the binary-LWE recently, we choose secret key from binary set and modify the basic encryption scheme proposed in Linder and Peikert in 2010. We propose a fully homomorphic encryption scheme based on the new basic encryption scheme. We analyze the correctness and give the proof of the security of our scheme. The public key, evaluation keys and tensored ciphertext have better size in our scheme. （2） Estimating parameters for fully homomorphic encryption scheme is an important work. We estimate the concert parameters for our scheme. We compare these parameters between our scheme and Bral2 scheme. Our scheme have public key and private key that smaller by a factor of about logq than in Bral2 scheme. Tensored ciphertext in our scheme is smaller by a factor of about log2q than in Bral2 scheme. Key switching matrix in our scheme is smaller by a factor of about log3q than in Bra12 scheme.
文摘同态加密(homomorphic encryption,HE)由于低执行效率和无法保护数据完整性的问题严重限制了其在实际应用中的部署,尤其是在对延迟有严格要求的场景中,为此,提出了一种新的HE来解决这些问题并增强通用性.为了解决执行效率低的问题,设计了多线程矩阵乘法(multithreaded matrix multiplication,MMM)算法.利用MMM算法,可以将加密任务拆解分配给多个线程并行执行,达到加速的目的.针对恶意服务器场景下的数据篡改问题,设计了一个可验证加密机制,利用非交互零知识证明(zk-SNARK)技术保护外包计算中密文的完整性.结合MMM算法,设计了一种高效的基于零知识证明的可验证全同态加密算法(verifiable fully homomorphic encryption based on zk-SNARKs,zk-VFHE).理论分析和实验结果表明,zk-VFHE比同类协议具有更快的执行速度和更高的安全性.
文摘分析以往格上基于身份的全同态加密方案,指出方案效率低的根本原因在于陷门生成和原像采样过程的复杂度过高,为此提出一种新的解决方案。先将新型陷门函数与对偶容错学习(LWE,learning with errors)算法有机结合,构造一种新的格上基于身份的加密方案;再利用特征向量方法转化为格上基于身份的全同态加密方案。对比分析表明,所提方案的陷门生成复杂度显著降低,原像采样复杂度约降低为以往方案的1/3,SIVP近似因子缩小为以往方案的1/m^(1/2)。在标准模型下,方案安全性归约至判定性LWE的难解性,并包含严格的安全性证明。
文摘本文提出一种基于公钥密码体制(Number Theory Research Unit,NTRU)选择明文攻击(Chosen Plaintext Attack,CPA)可证明安全的全同态加密方案.首先,对NTRU的密钥生成算法进行改进,通过格上的高斯抽象算法生成密钥对,避免了有效的格攻击,同时,没有改变密钥的分布.然后,基于改进的NTRU加密算法,利用Flattening技术,构造了一个全同态加密体制,并在标准模型下证明方案是选择明文攻击不可区分性IND-CPA安全的.
