【作者】 李楠；

【导师】 黄勤龙；

【作者基本信息】 北京邮电大学 ， 网络空间安全， 2021， 硕士

【摘要】 目前,云计算凭借可扩展性、节省成本、访问方便等特点被认为是实现用户间消息共享的有效工具,可以显著地提高生产效率。基于云计算技术,云计算平台能在数秒内处理数以千万甚至亿万的信息,这些信息会通过互联网提供给用户和厂商,从而实现资源受限的用户对软件的管理以及硬件的维护。云计算的迅猛发展促使各类以云计算为基础的应用场景应用而生,典型应用场景有数据协作和消息分发。在典型的协作方案中,数据所有者将数据外包给云平台,用户可以访问和共享数据。考虑到云平台的半可信,目前很多方案都选择使用基于属性加密(Attribute-Based Encryption,ABE)算法来保证数据机密性和细粒度的访问控制。但是,如何在灵活、动态的访问策略下仅允许授权用户正确访问协作数据是一个具有挑战性的问题。此外,云计算平台也被广泛用于用户和物联网设备的消息分发,物联云应运而生。借助物联云,移动端用户可以定期远程控制其智能设备,其中的群组消息分发服务对于物联云中的通信已变得不可缺少。同样地,目前ABE因其细粒度访问控制特性已经被用来保护物联云中消息的机密性。但是,只实现机密性是远远不够的,因为物联网设备易遭受攻击,发件人的真实性和前向保密性对于群组消息的传递非常有必要,即消息服务中的安全性和隐私性也是用户最关心的问题之一。对于上述提到的安全性问题,本文在进行大量的相关基础研究基础上,剖析了两个主要的关注方向并给出了解决方案:1)提出了一种基于可扩展属性加密的云数据安全共享机制。该方案在云计算场景中实现可扩展的访问控制和增强的完整性检查。本文首先介绍一个基于密文策略属性基加密的有效策略扩展框架,该框架允许满足当前访问策略的用户自定义新访问策略,并以非限制性或限制性方式将其添加到当前策略中;同时,本文实现了抵抗恶意用户的完整性检查,从而可确保所添加的访问策略来自授权用户。为了完善模型,本文进一步实现了抵抗半可信云服务器的完整性检查,借助基于排名的默克里哈希树实现动态公共审计,允许用户在解密之前验证密文的完整性,并支持在策略扩展阶段对密文更新操作进行有效的完整性标记更新。安全分析和实验结果表明,本文的方案对于云计算中的数据协作场景是安全高效的。2)提出了一种基于可穿刺双策略属性加密的云数据安全共享机制。本文首先给出了可穿刺双策略属性加密(Pt-DP-ABE)的框架,该框架允许发送者和接收者都可以指定其访问策略,并且合法用户可以对私钥进行穿刺以撤消被所选标签标记的密钥的解密能力。在此基础上,为了更好的完善模型,本文进一步将“对发送者身份的认证”特性纳入考虑,借助非交互式零知识证明系统,设计并实现物联云中基于属性的可认证群组消息传递(AB-AGMD)框架,该方案可保护群组消息的机密性、真实性及密文的前向保密性。最后,本文证明了AB-AGMD的安全性,并实施了仿真实验以证明它对于物联云中安全的群组消息传递是实用的。更多还原

【Abstract】 Nowadays,cloud computing is considered to be an effective tool to enable data sharing among different users due to its scalability,budget saving and convenient access,and can significantly improve production efficiency.Based on cloud computing technology,the cloud computing platform can process tens of millions or even hundreds of millions of information within a few seconds,and this information will be provided to users and manufacturers via the Internet,so that users with limited resources can manage software and maintain hardware.The rapid development of cloud computing has prompted the emergence of various cloud computing-based application scenarios.Typical application scenarios based on cloud computing include data collaboration and group messaging distribution.In a typical collaboration scenario,data owner outsources the data to cloud platform,and users can access and share the data.In consideration of the semi-trusted cloud platform,attribute-based encryption(ABE)has been utilized to guarantee data confidentiality and fine-grained access control.However,how to allow the collaborated data to be accessed only by authorized users correctly under flexible and dynamic access policy is a challenging problem.In addition,cloud computing platforms have also been widely used for message distribution between users and Internet of things(IoT)devices,and Cloud IoT has emerged.With the help of Cloud IoT,mobile users can regularly remotely control their smart devices.Group messaging services have become indispensable for communications in the IoT cloud.Currently,ABE has been used to protect the confidentiality of messages in the cloud IoT due to its fine-grained access control features.However,only achieving confidentiality is not enough,as IoT devices are vulnerable to attacks,so that authenticity and forward secrecy of the sender are particularly important for group messaging.Data security and privacy in messaging services are also one of the most concerned issues for users.For the security issues mentioned above,we analyzes the two main concerns and provides solutions based on a large number of relevant basic research:1)A cloud data security sharing scheme based on extendable attribute encryption is proposed.This scheme implements extendable access control and enhanced integrity checking in cloud computing.We first introduce an efficient policy extending framework with ciphertext-policy ABE,which allows users who satisfy the current access policy to customize new access policies and add them to current policies in a non-restrictive or restrictive way.At the same time,we achieve integrity checking against the malicious users with equality test algorithm,so that it can be ensured that the added access policy comes from an authorized user.In order to improve the model,we further provide a dynamic public auditing protocol based on rank-based Merkle Hash Tree,which permits users to verify the ciphertext integrity before decryption and supports efficient integrity tag updating for ciphertext updating operations in the policy extending phase.The security analysis and experimental results indicate that our scheme is secure and efficient for data collaboration scenario in cloud computing.2)A cloud data security sharing scheme based on puncturable dual-policy ABE is proposed.We first propose a puncturable dual-policy ABE(Pt-DP-ABE)scheme in which both senders and receivers are allowed to specify their policies,and legitimate users can puncture the private key to revoke the decryption ability of the key marked by the selected tag.On this basis,we further take into account the property of "authenticate the sender’s identity",with the aid of a non-interactive zero-knowledge proof system,design and implement attribute-based authenticated group message delivery(AB-AGMD)scheme in cloud IoT,which can protect the confidentiality and authenticity of group messages,and also ensures forward secrecy.Finally,we prove the security of AB-AGMD,and implement the experimental system to show that it is practical for secure group message delivery in cloud IoT.更多还原