In the driven of big data, social computing and information security is undergoing rapid development and beginning to cross. This paper describes a key-policy attribute-based signcryption scheme which has less computation costs than existing similar schemes by utilizing secure outsourcing of scientific computation in cloud computing and eliminates overhead for users, the ciphertext is short, compact, the correctness of transformation algorithm is verifiable. The decrease of ciphertext is 17 %. Additionally, new scheme remits the key escrow problem and is proven selective security in the standard model, it could be verified publicly, applied in mobile devices.
随机预言模型下安全的密码方案并不能保证在现实世界中的安全性,而构造标准模型下、基于格上困难问题,可抗合谋攻击的身份型代理重加密方案一直是密码学界的研究热点。通过采用陷门产生函数、左右抽样、bit分解等技术,设计了一个抗合谋攻击安全的基于身份代理重加密方案。在标准模型下,将该方案的安全性规约为判定性LWE(learning with errors)困难假设,证明了其能够达到IND-Pr ID-CPA安全,并对单向、多跳和抗合谋攻击等性质进行了分析。
With the development of quantum computer, multivariate public key cryptography withstanding quantum attack has became one of the research focus. The existed signcryption schemes from discrete logarithm and bilinear paring are facing the serious threats. Based on multivariate public key cryptography, a new certificateless multi-receiver hybrid signcryption scheme has been proposed. The proposal reduced the cipher text and could handle arbitrary length messages by employing randomness reusing and hybrid encryption, as well as keeping security. In the random oracle model, the scheme＇s confidentiality could withstand the IND-CCA2 adversary and its unforgeability could withstand the UF-CMA adversary under the hardness of multivariat quadratic （MQ） problem and isomorphism of polynomials （IP） assumption. It has less computation overhead and higher transmission efficiency than others. It reduced 33% cipher data compared with the existed similar scheme.
