混淆在可证明安全中的理论与应用研究

发布时间：2018-04-27 08:16

本文选题：混淆 + 平均情况下的虚拟黑盒特性　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：混淆（Obfuscation）问题近年来是密码学研究的热点和难点之一。混淆来源于程序混淆，最开始是为了保护软件的知识产权，防止软件遭到静态分析、篡改或逆向工程攻击。近年来经过Barak，Canetti，Goldwasser等密码学专家的研究和讨论，混淆已发展为特有的密码学问题。混淆的目的在于通过一定的技术手段使得原有程序变得“不可识别”，但同时保持原有程序的功能不变。目前，混淆在多方面取得了重要进步，比如说使用点函数（Point Function）构造的混淆、概率加密和混淆之间的联系、基于不同黑盒特性（Black-Box Property）的混淆安全性的分析、对普遍功能的混淆和特殊功能的混淆的可能性的讨论等。以上的结论使得混淆在密码学中的研究主要分为以下两个方面：一、讨论点函数和多比特输出点函数（Point Functions with Multibit Output）构造简单算法的抽象混淆的研究；二、讨论复杂算法协议和混淆的结合，构造特殊功能（Special Functionality）的安全混淆，比如说重加密、加密签名的安全混淆、功能性重加密等。本文的主要成果如下：一、介绍了代码混淆的概念、目的和几种方法，包括外形混淆、数据混淆、预防混淆、控制结构混淆等几种结构和具体方法；二、介绍了可证明安全的基本思想、困难问题假设、数字签名方案和公钥加密方案的安全模型，以及标准模型方法和随机预言机模型方法；三、介绍了密码学上的混淆的应用范围、三种特性，介绍了图灵机混淆器和回路混淆器的定义，并比较了两种混淆器的异同和相互关系；四、总结了已有的特殊功能的安全混淆，包括重加密安全混淆、加密签名安全混淆、功能性重加密和抗共谋混淆和它们的研究背景、具体算法和安全性分析；五、提出了一种将严格结构保留签名的特殊代理签名方案，在未知第三方是否可信的情况下，在代理签名过程中的密钥传输阶段对密钥进行安全混淆，成功地防止签名代理过程中密钥被窃听或者欺骗，保障了代理签名的顺利实施。我们在判定性线性（DL）假设和判定双线性Diffie-Hellman（DBDH）假设，并联系平均情况下的虚拟黑盒特性进行证明。这个混淆程序把单一用户的加密签名的安全混淆扩展到多用户设置的情况，并且它的安全性将不受密文数和用户数量的影响。它既保留了严格结构保留签名的私钥生成模式的优势和其高效的运算效率，，又在第三方可能不可信问题上得到一种解决方案，降低了第三方共谋攻击的威胁。
[Abstract]:Obfuscation (obfuscation) problem has been one of the hot and difficult problems in cryptography in recent years. Confusion originates from program confusion, which is to protect the intellectual property of software from static analysis, tampering or reverse engineering attack. In recent years, through the research and discussion of Baraki Canettii Goldwasser and other cryptographic experts, confusion has developed into a unique cryptographic problem. The purpose of confusion is to make the original program "unrecognizable" by certain technical means, but at the same time to keep the function of the original program unchanged. At present, obfuscation has made important progress in many aspects, such as the confusion constructed with point function, the relation between probabilistic encryption and obfuscation, and the analysis of the confusion security based on different black-box features (Black-Box property). Discussion of the possibility of confusion between universal function and special function. The above conclusions make the study of obfuscation in cryptography mainly divided into the following two aspects: first, discuss the abstract confusion of point function and multi-bit output point function Functions with Multibit Output) construction simple algorithm; second, This paper discusses the combination of complex algorithm protocol and confusion, and constructs the security confusion of special function, such as reencryption, security confusion of encryption signature, functional reencryption, etc. The main results of this paper are as follows: Firstly, this paper introduces the concept, purpose and several methods of code confusion, including shape confusion, data confusion, prevention of confusion, control structure confusion and so on. Secondly, the basic idea of provable security, the assumption of difficult problem, the security model of digital signature scheme and public key encryption scheme, the standard model method and random oracle model method are introduced. Thirdly, the application scope and three characteristics of cryptography confusion are introduced. The definitions of Turing machine mixer and loop mixer are introduced, and the similarities and differences between the two kinds of confusion are compared. 4. Summarize the existing security confusion of special functions, including reencryption security confusion, encryption signature security confusion, functional re-encryption and anti-conspiracy confusion and their research background, specific algorithms and security analysis; 5. A special proxy signature scheme is proposed, which preserves the signature with strict structure, and confuses the key safely in the key transmission stage of the proxy signature under the condition of whether the unknown third party is trusted or not. In the process of signature proxy, the key is eavesdropped or deceived, which ensures the smooth implementation of proxy signature. We prove the determinative linear DL hypothesis and the bilinear Diffie-Hellman DBDH hypothesis with the average case of virtual black box. This obfuscation program extends the security obfuscation of a single user's encrypted signature to the case of multi-user settings, and its security will not be affected by the number of ciphertext and the number of users. It not only preserves the advantages of the private key generation mode with strict structure and signature and its efficient operation efficiency, but also obtains a solution to the possible untrusted problem of the third party, which reduces the threat of the third party collusion attack.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN918.1

【相似文献】