面向车联网数据隐私保护的高效分布式模型共享策略

针对车联网隐私数据共享面临的效率问题,提出了基于区块链的高效分布式模型共享策略.针对车联网场景下多实体、多角色的数据共享需求,通过在车辆、路边单元和基站之间构建主从链架构,实现了分布式模型安全共享;提出了基于激励机制的异步联邦学习算法,以激励车辆及路边单元参与优化过程;构造了混合PBFT的改进DPoS共识算法来降低通信...     

基于代理重加密的区块链数据受控共享方案

区块链以分布式共享全局账本的形式存储交易数据，数据共享难以实现隐私保护和可用性之间的平衡，现有的区块链数据共享方案在进行隐私保护的同时可用性较低，有效实现区块链数据访问权限的动态调整是一个挑战性问题.为此，本文提出基于代理重加密的区块链数据受控共享方案.首先，基于SM2构造代理重加密算法，并借此设计区块链数据受控共享方案，利用代理重加密保护交易数据隐私实现数据安全共享.其次，提出用户权限动态调整机制，区块链节点分工代理并对重加密密钥参数分割管理，实现用户访问权限确定性更新，交易数据的可见性得到动态调整.最后，安全性和性能分析表明，本方案可以在保护交易隐私的同时，实现区块链数据动态共享，并且在计算开销方面具有优势，更好地适用于区块链数据受控共享.     

考虑属性加密的物联网隐私数据跨域安全共享模型

物联网隐私数据共享中，由于远距离的高速跨域问题，在数据共享时存在安全隐患，为此设计考虑属性加密的物联网隐私数据跨域安全共享模型。考虑属性加密方法设计重量级物联网隐私数据属性加密方案，实现数据加密模块的设计。基于区块链技术设计跨域安全共享模块设置三个实体，分别为监管中心、云存储器与联盟区块链，实现数据的跨域共享。在角色信誉评估模块中，考虑角色信誉评估角色的恶意行为实施其信誉评估，将评估结果作为选择数据共享对象时的数据基础。测试结果表明：所构建模型的通信开销较低；在4台普通计算机总跨域距离为200 km时，跨域数据共享时间仅为45 000 ms左右；在各阶段其操作数量与加解密时间都较低。     

适用于物联网数据共享的区块链节点存储优化方案

在基于区块链的物联网数据共享中,区块链要求节点具备大量的存储资源,这极大阻碍了更多物联网设备加入共享。针对这个问题,提出一种适用于物联网数据共享的区块链节点存储优化方案。该方案提出适用于物联网数据共享场景的存储模型,引入部分重复码对实用拜占庭容错算法进行改进,在共识过程中完成区块链账本的优化存储。实验分析表明,该方案能够在保证共识效率和容错性的同时,大幅降低区块链节点的存储开销。     

区块链上基于云辅助的密文策略属性基数据共享加密方案

针对云存储的集中化带来的数据安全和隐私保护问题,该文提出一种区块链上基于云辅助的密文策略属性基(CP-ABE)数据共享加密方案。该方案采用基于属性加密技术对加密数据文件的对称密钥进行加密,并上传到云服务器,实现了数据安全以及细粒度访问控制;采用可搜索加密技术对关键字进行加密,并将关键字密文上传到区块链(BC)中,由区块链进行关键字搜索保证了关键字密文的安全,有效地解决现有的云存储共享系统所存在的安全问题。该方案能够满足选择明文攻击下的不可区分性、陷门不可区分性和抗串联性。最后,通过性能评估,验证了该方案的有效性。     

关于车联网联邦学习中可信信誉管理方法的研究

联邦学习由于其分布式、隐私保护等特点有望应用到车联网中，然而由于缺少相应的本地模型质量验证机制，全局模型容易受到恶意用户的攻击从而导致模型训练的准确率降低。提出一种车联网中分层区块链使能的联邦学习信誉管理架构。首先介绍整个架构的组成以及具体的工作流程，然后设计智能合约为系统提供更加灵活可信的信誉意见共享环境，并开发一种轻量级的区块链共识算法，以提升区块链的运行效率。仿真结果表明所提方法能够筛选出恶意用户，同时保证数据隐私和安全，从而提高FL的准确性。     

区块链上基于云辅助的密文策略属性基数据共享加密方案

针对云存储的集中化带来的数据安全和隐私保护问题,该文提出一种区块链上基于云辅助的密文策略属性基(CP-ABE)数据共享加密方案.该方案采用基于属性加密技术对加密数据文件的对称密钥进行加密,并上传到云服务器,实现了数据安全以及细粒度访问控制;采用可搜索加密技术对关键字进行加密,并将关键字密文上传到区块链(BC)中,由区块链进行关键字搜索保证了关键字密文的安全,有效地解决现有的云存储共享系统所存在的安全问题.该方案能够满足选择明文攻击下的不可区分性、陷门不可区分性和抗串联性.最后,通过性能评估,验证了该方案的有效性.     

基于条件代理重加密的区块链医疗数据共享模型

医疗数据具有较高的私密性和敏感性，然而以往的医疗数据共享过程中往往采取中心化的集中式存储，这带来了数据隐私泄露和非法篡改的风险。针对上述问题，文中将区块链和代理重加密技术结合起来，提出基于条件代理重加密的区块链医疗数据共享模型。各级医院组成联盟区块链，采用改进的实用拜占庭共识机制选取代理节点，提高了节点达成共识的效率。借助于云服务器海量的存储能力，将患者的医疗数据存储在云服务器，联盟区块链中保存条件摘要密文，数据请求者在联盟链上进行条件搜索获取患者医疗数据，同时利用代理重加密技术实现特定数据请求者和第三方用户之间的细粒度共享。实验结果证明，基于条件代理重加密的区块链医疗数据共享模型在通信和计算开销方面更具优势。     

基于区块链的联邦学习应用研究

联邦学习区块链应用领域、架构特性和隐私机制等,具有很强的互补性和兼容性。将这两种技术结合起来,以增加隐私保护。数据交换是在计算性能的机制进行。基于区块链的联邦学习,了解有关开发基于区块链的联邦学习最新研究成果。人工智能离不开大数据,由于目前的数据管控政策和行业竞争造成的数据孤岛严重限制了大数据技术的使用价值。联邦学习可以消除数据孤岛,在多个参与者不公开数据集的情况下,共同完成模型学习。由于中心化的相互依赖,以及隐私泄露的风险,基于区块链的联邦学习方法已进入人工智能研究领域。基于此,本文讨论了联邦学习和区块链的概念,结合区块链和联邦学习进行了分析。     

区块链架构下高效的车联网跨域数据安全共享研究

为了解决车联网环境下跨信任域数据共享中跨域数据泄露严重、跨域共享不可控、跨域访问效率低的问题，提出了一种区块链架构下高效的车联网跨域数据安全共享方案。不同信任域的可信机构构成区块链，采用改进的密文策略属性基加密算法加密数据，结合区块链和星际文件系统进行存储，构建了基于区块链的跨域数据细粒度、安全共享方案；设计了基于混淆布隆过滤器的跨域访问验证方法，智能合约基于链上访问策略进行快速的解密测试，提高大量跨域密文的访问效率；设计了基于外包解密的跨域数据获取方法，可信机构为跨域访问请求进行密文转换，并执行包含复杂双线性配对运算的外包解密，减少了车辆在解密过程的时间开销。实验结果表明，所提方案有效提高了跨域密文转换和车辆解密的效率，与现有方案相比，跨域数据访问效率平均提升了60%。     

An intelligent and privacy-enhanced data sharing strategy for blockchain-empowered Internet of Things

With the development of the Internet of Things (IoT), the massive data sharing between IoT devices improves the Quality of Service (QoS) and user experience in various IoT applications. However, data sharing may cause serious privacy leakages to data providers. To address this problem, in this study, data sharing is realized through model sharing, based on which a secure data sharing mechanism, called BP2P-FL, is proposed using peer-to-peer federated learning with the privacy protection of data providers. In addition, by introducing the blockchain to the data sharing, every training process is recorded to ensure that data providers offer high-quality data. For further privacy protection, the differential privacy technology is used to disturb the global data sharing model. The experimental results show that BP2P-FL has high accuracy and feasibility in the data sharing of various IoT applications.     

数据新鲜度驱动的协作式无人机联邦学习智能决策优化研究

联邦学习是6G关键技术之一,其可以在保护数据隐私的前提下,利用跨设备的数据训练一个可用且安全的共享模型。然而,大部分终端设备由于处理能力有限,无法支持复杂的机器学习模型训练过程。在异构网络融合环境下移动边缘计算(MEC)框架中,多个无人机(UAVs)作为空中边缘服务器以协作的方式灵活地在目标区域内移动,并且及时收集新鲜数据进行联邦学习本地训练以确保数据学习的实时性。该文综合考虑数据新鲜程度、通信代价和模型质量等多个因素,对无人机飞行轨迹、与终端设备的通信决策以及无人机之间的协同工作方式进行综合优化。进一步,该文使用基于优先级的可分解多智能体深度强化学习算法解决多无人机联邦学习的连续在线决策问题,以实现高效的协作和控制。通过采用多个真实数据集进行仿真实验,仿真结果验证了所提出的算法在不同的数据分布以及快速变化的动态环境下都能取得优越的性能。     

A federated semi-supervised learning approach for network traffic classification

The classification of network traffic, which involves classifying and identifying the type of network traffic, is the most fundamental step to network service improvement and modern network management. Classic machine learning and deep learning methods have widely adopted in the field of network traffic classification. However, there are two major challenges in practice. One is the user privacy concern in cross-domain traffic data sharing for the purpose of training a global classification model, and the other is the difficulty to obtain large amount of labeled data for training. In this paper, we propose a novel approach using federated semi-supervised learning for network traffic classification, in which the federated server and clients from different domains work together to train a global classification model. Among them, unlabeled data are used on the client side, and labeled data are used on the server side. The experimental results derived from a public dataset show that the accuracy of the proposed approach can reach 97.81%, and the accuracy gap between the federated learning approach and the centralized training method is minimal.     

Federated learning-based privacy-preserving electricity load forecasting scheme in edge computing scenario

In the era of big data, massive amounts of data hold great value. However, much data exists as isolated islands, and the maximum value of the data cannot be fully utilized. Federated learning allows each client to train local data and then share the training model parameters securely, which can address the isolated data island problem and exploit data value while ensuring data privacy and security. Accordingly, in order to securely complete the electric power load forecasting using existing data, this paper constructs a federated learning-based privacy-preserving scheme to support electricity load forecasting in edge computing scenarios. To address the problems of the data-isolated islands and data privacy in electric power systems, this paper proposes a decentralized distributed solution based on the federated learning technique. Our scheme achieves electricity load forecasting for power systems through the federated learning-based framework and uses edge computing architecture to improve real-time data capability and reduce network latency. For the hierarchical scheduling structure in power systems, we divide the system into a cloud-side-device three-layer architecture, which achieves structural coordination and balance, and each layer collects information according to the scheduling control tasks, promoting scheduling effectiveness. Finally, different privacy protection methods are used on the cloud-edge and edge-device sides to significantly enhance data security. Moreover, We have conducted extensive experimental simulations for our proposed scheme. The experimental results show that the relative error of electricity load forecasting is around 1.580%. Meanwhile, our scheme achieves high accuracy and low memory usage. The security analysis proves the feasibility and security of our scheme.     

Towards asynchronous federated learning for heterogeneous edge-powered internet of things

The advancement of the Internet of Things (IoT) brings new opportunities for collecting real-time data and deploying machine learning models. Nonetheless, an individual IoT device may not have adequate computing resources to train and deploy an entire learning model. At the same time, transmitting continuous real-time data to a central server with high computing resource incurs enormous communication costs and raises issues in data security and privacy. Federated learning, a distributed machine learning framework, is a promising solution to train machine learning models with resource-limited devices and edge servers. Yet, the majority of existing works assume an impractically synchronous parameter update manner with homogeneous IoT nodes under stable communication connections. In this paper, we develop an asynchronous federated learning scheme to improve training efficiency for heterogeneous IoT devices under unstable communication network. Particularly, we formulate an asynchronous federated learning model and develop a lightweight node selection algorithm to carry out learning tasks effectively. The proposed algorithm iteratively selects heterogeneous IoT nodes to participate in the global learning aggregation while considering their local computing resource and communication condition. Extensive experimental results demonstrate that our proposed asynchronous federated learning scheme outperforms the state-of-the-art schemes in various settings on independent and identically distributed (i.i.d.) and non-i.i.d. data distribution.     

基于内积谓词的属性基隐私保护加密方案

隐私保护是信息安全中的热点话题,其中属性基加密(ABE)中的隐私问题可分为数据内容隐私、策略隐私及属性隐私。针对数据内容、策略和属性3方面隐私保护需求,该文提出基于内积谓词的属性基隐私保护加密方案(PPES)。所提方案利用加密算法的机密性保障数据内容隐私,并通过向量承诺协议构造策略属性及用户属性盲化方法,实现策略隐私及属性隐私。基于混合论证技术,该文证明了所提方案满足标准模型下适应性选择明文安全,且具备承诺不可伪造性。性能分析结果显示,与现有方法相比,所提方案具有更优的运行效率。     

Encrypted data sharing with multi-owner based on digital rights management in online social networks简

The online social networks（OSNs） offer attractive means for social interactions and data sharing, as well as raise a number of security and privacy issues. Although current solutions propose to encrypt data before sharing, the access control of encrypted data has become a challenging task. Moreover, multiple owners may enforce different access policy to the same data because of their different privacy concerns. A digital rights management（DRM） scheme is proposed for encrypted data in OSNs. In order to protect users＇ sensitive data, the scheme allows users outsource encrypted data to the OSNs service provider for sharing and customize the access policy of their data based on ciphertext-policy attribute-based encryption. Furthermore, the scheme presents a multiparty access control model based on identity-based broadcast encryption and ciphertext-policy attribute-based proxy re-encryption, which enables multiple owners, such as tagged users who appear in a single data, customize the access policy collaboratively, and also allows the disseminators update the access policy if their attributes satisfy the existing access policy. Security analysis and comparison indicate that the proposed scheme is secure and efficient.     

Hybrid cloud approach for block-level deduplication and searchable encryption in large universe

Ciphertext-policy attribute-based searchable encryption (CP-ABSE) can achieve fine-grained access control for data sharing and retrieval, and secure deduplication can save storage space by eliminating duplicate copies. However, there are seldom schemes supporting both searchable encryption and secure deduplication. In this paper, a large universe CP-ABSE scheme supporting secure block-level deduplication are proposed under a hybrid cloud mechanism. In the proposed scheme, after the ciphertext is inserted into bloom filter tree (BFT), private cloud can perform fine-grained deduplication efficiently by matching tags, and public cloud can search efficiently using homomorphic searchable method and keywords matching. Finally, the proposed scheme can achieve privacy under chosen distribution attacks block-level (PRV-CDA-B) secure deduplication and match-concealing (MC) searchable security. Compared with existing schemes, the proposed scheme has the advantage in supporting fine-grained access control, block-level deduplication and efficient search, simultaneously.     

FastProtector：一种支持梯度隐私保护的高效联邦学习方法

联邦学习存在来自梯度的参与方隐私泄露,现有基于同态加密的梯度保护方案产生较大时间开销且潜在参与方与聚合服务器合谋导致梯度外泄的风险,为此,该文提出一种新的联邦学习方法FastProtector,在采用同态加密保护参与方梯度时引入符号随机梯度下降(SignSGD)思想,利用梯度中正负的多数决定聚合结果也能使模型收敛的特性,量化梯度并改进梯度更新机制,降低梯度加密的开销;同时给出一种加性秘密共享方案保护梯度密文以抵抗恶意聚合服务器和参与方之间共谋攻击;在MNIST和CIFAR-10数据集上进行了实验,结果表明所提方法在降低80%左右加解密总时间的同时仍可保证较高的模型准确率。     

基于相似度聚类的可信联邦安全聚合算法

联邦学习能够有效地规避参与方数据隐私问题,但模型训练中传递的参数或者梯度仍有可能泄露参与方的隐私数据,而恶意参与方的存在则会严重影响聚合过程和模型质量。基于此,该文提出一种基于相似度聚类的可信联邦安全聚合方法(FSA-SC)。首先基于客户端训练数据集规模及其与服务器间的通信距离综合评估选出拟参与模型聚合的候选客户端;然后根据候选客户端间的相似度,利用聚类将候选客户端划分为良性客户端和异常客户端;最后,对异常客户端类中的成员利用类内广播和二次协商进行参数替换和记录,检测识别恶意客户端。为了验证FSA-SC的有效性,以联邦推荐为应用场景,选取MovieLens 1M,Netflix数据集和Amazon抽样数据集为实验数据集,实验结果表明,所提方法能够实现高效的安全聚合,且相较对比方法有更高的鲁棒性。