Privacy-preserving Requirements Research Articles

DB-GPT: Large Language Model Meets Database

Large language models (LLMs) have shown superior performance in various areas. And LLMs have the potential to revolutionize data management by serving as the "brain" of next-generation database systems. However, there are several challenges that utilize LLMs to optimize databases. First, it is challenging to provide appropriate prompts (e.g., instructions and demonstration examples) to enable LLMs to understand the database optimization problems. Second, LLMs only capture the logical database characters (e.g., SQL semantics) but are not aware of physical characters (e.g., data distributions), and it requires to fine-tune LLMs to capture both physical and logical information. Third, LLMs are not well trained for databases with strict constraints (e.g., query plan equivalence) and privacy-preserving requirements, and it is challenging to train database-specific LLMs while ensuring database privacy. To overcome these challenges, this vision paper proposes a LLM-based database framework (DB-GPT), including automatic prompt generation, DB-specific model fine-tuning, and DB-specific model design and pre-training. Preliminary experiments show that DB-GPT achieves relatively good performance in database tasks like query rewrite and index tuning. The source code and datasets are available at github.com/TsinghuaDatabaseGroup/DB-GPT.

Multiparty Secure Broad Learning System for Privacy Preserving.

Multiparty learning is an indispensable technique to improve the learning performance via integrating data from multiple parties. Unfortunately, directly integrating multiparty data could not meet the privacy-preserving requirements, which then induces the development of privacy-preserving machine learning (PPML), a key research task in multiparty learning. Despite this, the existing PPML methods generally cannot simultaneously meet multiple requirements, such as security, accuracy, efficiency, and application scope. To deal with the aforementioned problems, in this article, we present a new PPML method based on the secure multiparty interactive protocol, namely, the multiparty secure broad learning system (MSBLS) and derive its security analysis. To be specific, the proposed method employs the interactive protocol and random mapping to generate the mapped features of data, and then uses efficient broad learning to train the neural network classifier. To the best of our knowledge, this is the first attempt for privacy computing method that jointly combines secure multiparty computing and neural network. Theoretically, this method can ensure that the accuracy of the model will not be reduced due to encryption, and the calculation speed is very fast. Three classical datasets are adopted to verify our conclusion.

Privacy-Preserving Operation Management of Battery Swapping and Charging System With Dual-Based Benders Decomposition

The battery swapping and charging system (BSCS) is an emerging and promising infrastructure to provide the energy refueling service for EVs. However, each operator (i.e., battery swapping module (BSM) operator and battery charging module (BCM) operator) in BSCS has privacy-preserving requirement and individual interest, which brings challenge to the operation management of BSCS. To this end, this paper investigates the privacy-preserving operation management problem (OMP) of BSCS from a social perspective, in which a comprehensive battery-swapping-charging process is modeled. Specifically, we first model the OMP as a constrained mixed-integer programming (MIP), where the evolutionary of the dynamic battery inventory and the SoC of charging batteries are considered simultaneously. Due to the privacy-preserving requirements of different operators and the strong coupling between continuous and binary decision variables, a novel dual-based Benders decomposition (DBD) algorithm is then developed, which has the following two merits: 1) Each BCM and BSM operator can independently solve its own charging and swapping scheduling problem, thus preserving the privacy of individual operator; 2) The combination of dual decomposition and Benders decomposition techniques facilitates a parallel implementation, thus improving the scheduling efficiency and saving the computational time. Finally, simulation results are provided to validate the effectiveness and scalability of the proposed algorithm.

Privacy-preserving Resilient Consensus for Multi-agent Systems in a General Topology Structure

Recent advances of consensus control have made it significant in multi-agent systems such as in distributed machine learning, distributed multi-vehicle cooperative systems. However, during its application it is crucial to achieve resilience and privacy; specifically, when there are adversary/faulty nodes in a general topology structure, normal agents can also reach consensus while keeping their actual states unobserved. In this article, we modify the state-of-the-art Q-consensus algorithm by introducing predefined noise or well-designed cryptography to guarantee the privacy of each agent state. In the former case, we add specified noise on agent state before it is transmitted to the neighbors and then gradually decrease the value of noise so the exact agent state cannot be evaluated. In the latter one, the Paillier cryptosystem is applied for reconstructing reward function in two consecutive interactions between each pair of neighboring agents. Therefore, multi-agent privacy-preserving resilient consensus (MAPPRC) can be achieved in a general topology structure. Moreover, in the modified version, we reconstruct reward function and credibility function so both convergence rate and stability of the system are improved. The simulation results indicate the algorithms’ tolerance for constant and/or persistent faulty agents as well as their protection of privacy. Compared with the previous studies that consider both resilience and privacy-preserving requirements, the proposed algorithms in this article greatly relax the topological conditions. At the end of the article, to verify the effectiveness of the proposed algorithms, we conduct two sets of experiments, i.e., a smart-car hardware platform consisting of four vehicles and a distributed machine learning platform containing 10 workers and a server.

An Integrated Design of Heat and Power Market for Energy Hubs Considering the Security Constraints of the System

According to the development of Combined Heat and Power (CHP) units, and energy converter equipment such as Heat Pumps (HP), energy scheduling of thermal and electrical systems became more dependent on each other. Multi-energy systems with new technologies require new designs as Integrated Heat and Power Markets (IHPM) to use the complementary capabilities of multi-energy systems such as Energy Hubs (EH). In this regard, a decentralized market structure is provided to make the opportunity of trading heat and power bilaterally for connected EHs of the distribution system under an integrated market framework. The proposed market structure based on the ADMM improves the social welfare value, reduces the operating costs, and considers the privacy-preserving requirements for the market clearing process. Results analysis indicates that P2P energy trading enables the use of complementary capabilities of EHs, and also reduces the required reserve margin of each entity. According to the obtained results, the evolution of the IHPM would increase the line loading of the power system by up to 28%, which can lead to line congestion if the security constraints of the system not be considered in the electrical market design. The convergency and optimality of results are validated by a centralized model.

Privacy-Preserving Platooning Control of Vehicular Cyber–Physical Systems With Saturated Inputs

Metaverse allows the physical reality to tightly integrate with the digital universe. As one typical metaverse application, platooning control of vehicular cyber–physical systems has attracted extensive attention as it is beneficial to improve traffic efficiency, driving safety, and emission reduction. However, due to the open nature of wireless communication networks, the transmitted vehicle-to-vehicle (V2V) data packets become exposed to the public and concomitant data leakage can lead to unintended consequences to vehicular platoons. This article is concerned with the privacy-preserving platooning control issue of vehicular cyber–physical systems with input saturations. First, a novel distributed proportional-integral observer is proposed to estimate the full state of each vehicle, where the integral terms with a forgetting factor facilitate to realize the tradeoff between transient performance and steady-state performance for the platoon. Second, sampled-data-based dynamic encryption and decryption schemes, featuring a dynamic private key, are developed such that the encrypted and decrypted V2V data can be kept private to each platoon vehicle. It is then shown that the platooning control problem over a generic communication topology can be cast into the stability issue of an auxiliary dynamic system. Furthermore, sufficient conditions on the existence of the desired observer and controller gains as well as the private key parameter selection are derived to guarantee the desired platoon stability and privacy preservation requirements. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed control method.

Accelerating Distributed Optimization via Fixed-Time Convergent Flows

Distributed optimization has gained significant attention in recent years, primarily fueled by the availability of a large amount of data and privacy-preserving requirements. This paper presents a fixed-time convergent optimization algorithm for solving a potentially non-convex optimization problem using a first-order multi-agent system. Each agent in the network can access only its private objective function, while local information exchange is permitted between the neighbors. The proposed optimization algorithm combines a fixed-time convergent distributed parameter estimation scheme with a fixed-time distributed consensus scheme as its solution methodology. The results are presented under the assumption that the team objective function is strongly convex, as opposed to the common assumptions in the literature requiring each of the local objective functions to be strongly convex. The results extend to the class of possibly non-convex team objective functions satisfying only the Polyak-Lojasiewicz (PL) inequality. It is also shown that the proposed continuous-time scheme, when discretized using Euler's method, leads to consistent discretization.

Binarizing Split Learning for Data Privacy Enhancement and Computation Reduction

Split learning (SL) enables data privacy preservation by allowing clients to collaboratively train a deep learning model with the server without sharing raw data. However, SL still has limitations such as potential data privacy leakage and high computation for clients. In this paper, we propose to binarize the SL local layers for faster computation (up to 17.5 times less forward-propagation time in both training and inference phases on mobile devices) and reduced memory usage (up to 32 times less memory and bandwidth requirements). More importantly, the binarized SL (B-SL) model can reduce privacy leakage from SL smashed data with merely a small degradation in model accuracy. To further enhance privacy preservation, we also propose two novel approaches: 1) training with additional local leak loss and 2) applying differential privacy, which could be integrated separately or concurrently into the B-SL model. Experimental results with different datasets have affirmed the benefits of the B-SL models compared with several benchmark models. The effectiveness of B-SL models against feature-space hijacking attack (FSHA) is also illustrated. Our results have demonstrated B-SL models are promising for lightweight IoT/mobile applications with high privacy-preservation requirements such as mobile healthcare applications.

SecCDV: A Security Reference Architecture for Cybertwin-Driven 6G V2X

Recently, the proposed Cybertwin-driven 6G network architecture has attracted attentions from academia, which can support the communication anchor, security agent and other functions for physical devices. Especially, Cybertwin can provide promising Vehicle-to-Everything (V2X) applications in future. However, security and privacy of Cybertwin still face various threats and challenges, which may impact future deployments for Cybertwin-driven 6G V2X. In this paper, we first introduce the architecture and promising applications of Cybertwin-driven 6G V2X. Then, we analyze essential data security and privacy preservation requirements for Cybertwin-driven 6G V2X. Particularly, we present the security reference architecture of Cybertwin-driven 6G V2X. As a case study, we investigate the migration of Cybertwin caused by vehicle mobility, and propose a handover authentication scheme to create new Cybertwin between vehicle and edge server based on proxy ring signature technique. Finally, we discuss several open research directions for achieving more secure Cybertwin-driven 6G V2X.

Blockchain for Transparent Data Management Toward 6G

The wealth of user data acts as a fuel for network intelligence toward the sixth generation wireless networks (6G). Due to data heterogeneity and dynamics, decentralized data management (DM) is desirable for achieving transparent data operations across network domains, and blockchain can be a promising solution. However, the increasing data volume and stringent data privacy-preservation requirements in 6G bring significantly technical challenge to balance transparency, efficiency, and privacy requirements in decentralized blockchain-based DM. In this paper, we investigate blockchain solutions to address the challenge. First, we explore the consensus protocols and scalability mechanisms in blockchains and discuss the roles of DM stakeholders in blockchain architectures. Second, we investigate the authentication and authorization requirements for DM stakeholders. Third, we categorize DM privacy requirements and study blockchain-based mechanisms for collaborative data processing. Subsequently, we present research issues and potential solutions for blockchain-based DM toward 6G from these three perspectives. Finally, we conclude this paper and discuss future research directions.

Knowledge transfer-based distributed differential evolution for dynamic database fragmentation

Database fragmentation can protect the distributed database’s privacy by dividing attributes of sensitive associations into different fragments. Previous database fragmentation algorithms are designed for the initialization of the distributed database. However, the initial database fragmentation cannot maintain its effect during the distributed database’s entire life cycle. This paper defines a dynamic database fragmentation problem with privacy preservation requirements, in which both the privacy preservation degree and the communication cost are considered during the optimization. For this problem, a knowledge transfer-based distributed differential evolution algorithm (KT-DDE) is proposed to achieve the optimal communication cost and maintain privacy preservation. The proposed KT-DDE algorithm includes a distributed framework and a differential evolution-based optimizer. In the proposed distributed framework, the fragmentation knowledge is transferred between different database fragmentation subproblems. The fragmentation information of various individuals is exchanged in the optimizer and used to generate trial individuals. After the selection, competitive trial individuals are kept in the population. Experimental results show that the proposed algorithm can outperform the other competitors in terms of solution accuracy, convergence speed, and computation efficiency. In addition, the effectiveness of the proposed components is verified.

Memristor-Based Variation-Enabled Differentially Private Learning Systems for Edge Computing in IoT

Edge artificial intelligence (AI) achieves real-time local data analysis for IoT systems, enabling low-power and high-speed operation, but comes with privacy-preserving requirements. The memristor-based computing system is a promising solution for edge AI, but it needs a low-cost privacy protection mechanism due to limited resources. In this article, we propose a noise distribution normalization (NDN) method to add Gaussian distributed noise through hardware implementation, thereby achieving differential privacy in edge AI. Instead of using traditional algorithmic noise-insertion methods, we take advantage of inherent cycle-to-cycle variations of memristors during the weight-update process as the noise source, which does not incur extra software or hardware overhead. In one case study, the proposed method realizes ultralow-cost differentially private stochastic gradient descent (DP-SGD) for edge AI in IoT systems, achieving a 3.5%–15.5% average recognition accuracy improvement under different noise levels, as compared with a baseline mechanism.

The effectiveness of backward contact tracing in networks.

Effective control of an epidemic relies on the rapid discovery and isolation of infected individuals. Because many infectious diseases spread through interaction, contact tracing is widely used to facilitate case discovery and control. However, what determines the efficacy of contact tracing has not been fully understood. Here we reveal that, compared with 'forward' tracing (tracing to whom disease spreads), 'backward' tracing (tracing from whom disease spreads) is profoundly more effective. The effectiveness of backward tracing is due to simple but overlooked biases arising from the heterogeneity in contacts. We argue that, even if the directionality of infection is unknown, it is possible to perform backward-aiming contact tracing. Using simulations on both synthetic and high-resolution empirical contact datasets, we show that strategically executed contact tracing can prevent a substantial fraction of transmissions with a higher efficiency-in terms of prevented cases per isolation-than case isolation alone. Our results call for a revision of current contact-tracing strategies so that they leverage all forms of bias. It is particularly crucial that we incorporate backward and deep tracing in a digital context while adhering to the privacy-preserving requirements of these new platforms.

Data Management for Future Wireless Networks: Architecture, Privacy Preservation, and Regulation

Next-generation wireless networks (NGWN) aim to support diversified smart applications that require frequent data exchanges and collaborative data processing among multiple stakeholders. Data management (DM), including data collection, storage, sharing, and computation, plays an essential role in empowering NGWN. However, DM for NGWN faces two significant challenges: stakeholders' data cannot be easily managed across different trust domains under a distributed network architecture; and privacy preservation requirements of personal data become more rigorous under new privacy regulations. To explore possible solutions to address the challenges, we first investigate the state-of-the-art architecture designs for DM and emphasize advantages of a blockchain-based DM architecture. Then we summarize existing privacy-preserving techniques in terms of advantages and challenges when being applied to DM. In addition, we review recent privacy regulations with their impacts on DM and discuss the existing solutions with privacy regulation compliance based on blockchain. Finally, we identify further research directions for achieving DM with privacy preservation.

Guaranteed distributed machine learning: Privacy-preserving empirical risk minimization.

Distributed learning over data from sensor-based networks has been adopted to collaboratively train models on these sensitive data without privacy leakages. We present a distributed learning framework that involves the integration of secure multi-party computation and differential privacy. In our differential privacy method, we explore the potential of output perturbation and gradient perturbation and also progress with the cutting-edge methods of both techniques in the distributed learning domain. In our proposed multi-scheme output perturbation algorithm (MS-OP), data owners combine their local classifiers within a secure multi-party computation and later inject an appreciable amount of statistical noise into the model before they are revealed. In our Adaptive Iterative gradient perturbation (MS-GP) method, data providers collaboratively train a global model. During each iteration, the data owners aggregate their locally trained models within the secure multi-party domain. Since the conversion of differentially private algorithms are often naive, we improve on the method by a meticulous calibration of the privacy budget for each iteration. As the parameters of the model approach the optimal values, gradients are decreased and therefore require accurate measurement. We, therefore, add a fundamental line-search capability to enable our MS-GP algorithm to decide exactly when a more accurate measurement of the gradient is indispensable. Validation of our models on three (3) real-world datasets shows that our algorithm possesses a sustainable competitive advantage over the existing cutting-edge privacy-preserving requirements in the distributed setting.

A Search Efficient Privacy-Preserving Location-Sharing Scheme in Mobile Online Social Networks

With the advent of intelligent handheld devices, location sharing becomes one of the most popular services in mobile online social networks (mOSNs). In location-sharing services, users can enjoy a better social experience by updating their real-time location information. However, the leakage of private information may hinder the further development of location-sharing services. Although many solutions have been proposed to protect users’ privacy, the privacy-utility trade-offs must be considered. Therefore, we propose a new scheme called search efficient privacy-preserving location-sharing (SELS) system. In our scheme, we create a new approach named associated grids to improve the efficiency of location-sharing systems while maintaining users’ privacy. In addition, by setting the user-defined access control policy proposed in our scheme, users’ flexible privacy-preserving requirements can be satisfied. Detailed complexity and security analysis show that the proposed scheme is a practical and efficient privacy-preserving solution. Extensive simulations are performed to validate the effectiveness and performance of our scheme.

Umbra: A Visual Analysis Approach for Defense Construction Against Inference Attacks on Sensitive Information.

Collecting and analyzing anonymous personal information is required as a part of data analysis processes, such as medical diagnosis and restaurant recommendation. Such data should ostensibly be stored so that specific individual information cannot be disclosed. Unfortunately, inference attacks-integrating background knowledge and intelligent models-hinder classic sanitization techniques like syntactic anonymity and differential privacy from exhaustively protecting sensitive information. As a solution, we introduce a three-stage approach empowered within a visual interface, which depicts underlying inference behaviors via a Bayesian Network and supports a customized defense against inference attacks from unknown adversaries. In particular, our approach visually explains the process details of the underlying privacy preserving models, allowing users to verify if the results sufficiently satisfy the requirements of privacy preservation. We demonstrate the effectiveness of our approach through two case studies and expert reviews.

Blockchain Enables Your Bill Safer

As one of the most frequently used Internet-of-Things (IoT) devices, energy smart meter has been widely adopted to facilitate the measures of residential energy use. Residents pay for the bills from energy suppliers according to their monthly/seasonal usage. Practically, there is a demand from residents/governments to check whether the bills are in line with their real consumptions. However, it is challenging to realize this demand due to two critical problems. The first problem refers to the nonrepudiated privacy issue caused by access to residents’ energy consumption history (e.g., data integrity may be questioned, and residents’ daily timetables may be exposed). The second problem comes from the efficiency requirement for bulk auditing requests on residents’ bills and consumptions, usually risen by governments. So far, we have not found any solutions that can be directly used in this case. In this article, we propose using homomorphic encryption cooperated with the blockchain technique to leverage the data auditing and privacy-preserving requirements. We also employe a certificateless signature to resolve the efficiency bottleneck in batch auditing. This framework, called <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">pAuditChain</i> , not only accepts personal requests from residents for consumption checking but also handles bulk auditing requests issued by governments. To validate the correctness of the framework functions, we carried out a series of theoretical analysis, especially on the privacy preserving and auditing processes. To the best of our knowledge, the proposed framework is among the first solutions to improve the security and privacy of bills without losing the auditing function. Our approach concerns with IoT smart meters in energy supply industries and could be further extended to other forms of IoT devices with the bill demands.

EliMFS: Achieving Efficient, Leakage-Resilient, and Multi-Keyword Fuzzy Search on Encrypted Cloud Data

Motivated by privacy preservation requirements for outsourced data, keyword searches over encrypted cloud data have become a hot topic. Compared to single-keyword exact searches, multi-keyword fuzzy search schemes attract more attention because of their improvements in search accuracy, typo tolerance, and user experience in general. However, existing multi-keyword fuzzy search solutions are not sufficiently efficient when the file set in the cloud is large. To address this, we propose an Efficient Leakage-resilient Multi-keyword Fuzzy Search (EliMFS) framework over encrypted cloud data. In this framework, a novel two-stage index structure is exploited to ensure that search time is independent of file set size. The multi-keyword fuzzy search function is achieved through a delicate design based on the Gram Counting Order, the Bloom filter, and the Locality-Sensitive Hashing. Furthermore, considering the leakages caused by the two-stage index structure, we propose two specific schemes to resist these potential attacks in different threat models. Extensive analysis and experiments show that our schemes are highly efficient and leakage-resilient.

Spatial-temporal data-driven service recommendation with privacy-preservation

The ever-increasing popularity of web service sharing communities have produced a considerable amount of web services that share similar functionalities but vary in Quality of Services (QoS) performances. To alleviate the heavy service selection burden on users, lightweight recommendation ideas, e.g., Collaborative Filtering (CF) have been developed to aid users to select their preferred services. However, existing CF methods often face two challenges. First, service QoS is often context-aware and hence depends on the spatial and temporal information of service invocations heavily. While it requires challenging efforts to integrate both spatial and temporal information into service recommendation decision-making process simultaneously. Second, the location-aware and time-aware QoS data often contain partial sensitive information of users, which raise an emergent privacy-preservation requirement when performing service recommendations. In view of above two challenges, in this paper, we integrate the spatial-temporal information of QoS data and Locality-Sensitive Hashing (LSH) into recommendation domain and bring forth a location-aware and time-aware recommendation approach considering privacy concerns. At last, a set of experiments conducted on well-known WS-DREAM dataset show the feasibility of our approach.