On-device deep learning (DL) has emerged as a popular approach for mobile apps to deliver artificial intelligence services. Unlike traditional cloud-based approaches, it processes sensitive information locally, addressing severe concerns over sensitive data collection on cloud servers. However, this approach inevitably stores models on user devices and opens a new attack surface, i.e., adversarial weight attacks, which steer DL models to undesirable behaviors through direct model weight modification. Unfortunately, such risks stemming from on-device DL have been left unexplored. In this paper, we present the first practical adversarial weight attack against on-device DL models. To demonstrate this novel attack, we propose TYPHON, an automated attack system that removes the read-only restriction of on-device DL models through the reconstruction of writable counterparts and leverages the inference-only nature of on-device DL models to solve malicious parameters and manipulate model behaviors. Extensive experimental results across diverse datasets and model architectures confirm the superiority of our attack across multiple evaluation metrics. In addition, three real-world case studies are conducted with 100% attack success rates, demonstrating the practicality of TYPHON.

The critical energy infrastructure is undergoing two significant transformations: the rapid increase in renewable distributed energy resources (DER) and the digitalization of the energy sector, collectively shaping what is known as the Internet of Energy (IoE). Artificial intelligence (AI) has become a widely adopted tool for effectively allocating energy and managing sector-related resources, where ensuring fairness is essential. While inherent unfairness in AI systems is well acknowledged, little attention has been given to evaluating this unfairness and its real-world implications within the context of the IoE. In this study, we take a first step to elucidate the unfairness in AI-powered IoE systems induced by malicious users. We introduce Unfairness Score (UScore), a novel metric designed to evaluate the unfairness of machine learning models in real-world IoE scenarios. We then extensively evaluate unfairness attacks using three IoE tabular datasets, demonstrating that AI model fairness can be compromised through data poisoning, whether in centralized learning (CL) or federated learning (FL) settings. Notably, such compromises can occur when malicious users tamper with only a small subset of the data they control. Finally, we propose a novel approach that unifies fairness and differential privacy (DP) by leveraging DP as a provable defense mechanism. This approach provides a universally applicable solution to unfairness attacks, regardless of whether the learning tasks are classification or regression, and is effective in both FL and CL settings. Our contributions represent a significant step in addressing unfairness and privacy concerns in AI-powered IoE systems.

As encrypted traffic grows, traditional rule-based and deep learning methods struggle with engineering costs and encryption complexity. While Large Language Models (LLMs) offer promise for traffic analysis via pre-trained feature learning, they face challenges in handling diverse tasks, retaining pre-training knowledge, and adapting efficiently. To address these issues, we propose a new traffic representation learning method and a new Parameter-Efficient Fine-Tuning (PEFT) method for multi-task encrypted traffic analysis services, calledTrafficLLM.TrafficLLMalleviates task heterogeneity by utilizing a universal multi-task prompt template and addresses pre-training knowledge forgetting by integrating Singular Value Decomposition based Low-Rank Adaptation (SVD-LoRA). To further reduce the cost of adapting to multiple tasks, we combine the strengths of the Mixture of Experts (MoE) for multi-task learning with SVD-LoRA for PEFT, enabling efficient multi-task traffic analysis. Additionally, we introduce task-aware gating functions to dynamically assign different weights to experts, facilitating the efficient fusion of expert knowledge. Comprehensive experiments on 7 datasets across 5 downstream tasks demonstrate thatTrafficLLMdelivers superior analysis performance and resource efficiency compared to state-of-the-art models, including DeepSeek, NetGPT, ET-BERT, and TFE-GNN. Detailed analysis of throughput, memory usage, and latency further highlights the practical advantages ofTrafficLLM. Our data and code are available athttps://github.com/yiliucs/TrafficLLMhttps://github.com/yiliucs/TrafficLLM.

Software vulnerabilities (SVs) have emerged as a prevalent and crucial concern for safety-critical systems. This has spurred significant advancements in utilizing AI-based methods, including machine learning and deep learning, for software vulnerability detection (SVD). While AI-based methods have shown promising performance in SVD, their effectiveness on real-world, complex, and diverse source code datasets remains limited in practice. To tackle this challenge, in this paper, we propose a novel framework that enhances the capability of large language models to learn and utilize semantic and syntactic relationships from source code data for SVD. As a result, our proposed SAFE approach can enable the acquisition of fundamental knowledge from source code data while adeptly utilizing crucial relationships, i.e., semantic and syntactic associations, to improve the effectiveness of solving the SVD problem. The rigorous and extensive experimental results on three real-world challenging datasets (i.e., Devign, ReVeal, and D2A) demonstrate the superiority of our approach over eight effective and state-of-the-art baselines. In summary, on average, our SAFE approach achieves higher performances from 4.79% to 11.57% for F1-measure and from 16.93% to 26.24% for Recall compared to the baseline methods across all the datasets used.

Content providers increasingly utilise Content Delivery Networks (CDNs) to enhance users' content download experience. However, this deployment scenario raises significant security concerns regarding content confidentiality and user privacy due to the involvement of third-party providers. Prior proposals using private information retrieval (PIR) and oblivious RAM (ORAM) have proven impractical due to high computation and communication costs, as well as integration challenges within distributed CDN architectures. In response, we present \textsf{OblivCDN}, a practical privacy-preserving system meticulously designed for seamless integration with the existing real-world Internet-CDN infrastructure. Our design strategically adapts Range ORAM primitives to optimise memory and disk seeks when accessing contiguous blocks of CDN content, both at the origin and edge servers, while preserving both content confidentiality and user access pattern hiding features. Also, we carefully customise several oblivious building blocks that integrate the distributed trust model into the ORAM client, thereby eliminating the computational bottleneck in the origin server and reducing communication costs between the origin server and edge servers. Moreover, the newly-designed ORAM client also eliminates the need for trusted hardware on edge servers, and thus significantly ameliorates the compatibility towards networks with massive legacy devices.In real-world streaming evaluations, OblivCDN} demonstrates remarkable performance, downloading a $256$ MB video in just $5.6$ seconds. This achievement represents a speedup of $90\times$ compared to a strawman approach (direct ORAM adoption) and a $366\times$ improvement over the prior art, OblivP2P.

With large AI systems and models (LAMs) playing an ever-growing role across diverse applications, their impact on the privacy and cybersecurity of critical infrastructure has become a pressing concern. The LAMPS workshop is dedicated to tackling these emerging challenges, promoting dialogue on cutting-edge developments and ethical issues in safeguarding LAMs within critical infrastructure contexts. Bringing together leading experts from around the world, this workshop will delve into the complex privacy and cybersecurity risks posed by LAMs in critical sectors. Attendees will explore innovative solutions, exchange best practices, and contribute to shaping the future research agenda, emphasizing the crucial balance between advancing AI technologies and securing critical digital and physical infrastructures.

Phishing attacks remain a significant challenge for detection, explanation, and defense, despite over a decade of research on both technical and non-technical solutions. AI-based phishing detection methods are among the most effective approaches for defeating phishing attacks, providing predictions on the vulnerability label (i.e., phishing or benign) of data. However, they often lack intrinsic explainability, failing to identify the specific information that triggers the classification. To this end, we propose AI2TALE, an innovative deep learning-based approach for email (the most common phishing medium) phishing attack localization. Our method aims to not only predict the vulnerability label of the email data but also provide the capability to automatically learn and identify the most important and phishing-relevant information (i.e., sentences) in the phishing email data, offering useful and concise explanations for the identified vulnerability. Extensive experiments on seven diverse real-world email datasets demonstrate the capability and effectiveness of our method in selecting crucial information, enabling accurate detection and offering useful and concise explanations (via the most important and phishing-relevant information triggering the classification) for the vulnerability of phishing emails. Notably, our approach outperforms state-of-the-art baselines by 1.5% to 3.5% on average in Label-Accuracy and Cognitive-True-Positive metrics under a weakly supervised setting, where only vulnerability labels are used without requiring ground truth phishing information.

Graph neural networks (GNNs) have exhibited superior performance in various classification tasks on graph-structured data. However, they encounter the potential vulnerability from the link stealing attacks, which can infer the presence of a link between two nodes via measuring the similarity of its incident nodes' prediction vectors produced by a GNN model. Such attacks pose severe security and privacy threats to the training graph used in GNN models. In this work, we propose a novel solution, called Graph Link Disguise (GRID), to defend against link stealing attacks with the formal guarantee of GNN model utility for retaining prediction accuracy. The key idea of GRID is to add carefully crafted noises to the nodes' prediction vectors for disguising adjacent nodes as n-hop indirect neighboring nodes. We take into account the graph topology and select only a subset of nodes (called core nodes) covering all links for adding noises, which can avert the noises offset and have the further advantages of reducing both the distortion loss and the computation cost. Our crafted noises can ensure 1) the noisy prediction vectors of any two adjacent nodes have their similarity level like that of two non-adjacent nodes and 2) the model prediction is unchanged to ensure zero utility loss. Extensive experiments on five datasets are conducted to show the effectiveness of our proposed GRID solution against different representative link-stealing attacks under transductive settings and inductive settings respectively, as well as two influence-based attacks. Meanwhile, it achieves a much better privacy-utility trade-off than existing methods when extended to GNNs.

Dynamic graph neural networks (DGNNs) have emerged and been widely deployed in various web applications (e.g., Reddit) to serve users (e.g., personalized content delivery) due to their remarkable ability to learn from complex and dynamic user interaction data. Despite benefiting from high-quality services, users have raised privacy concerns, such as misuse of personal data (e.g., dynamic user-user/item interaction) for model training, requiring DGNNs to "forget" their data to meet AI governance laws (e.g., the "right to be forgotten" in GDPR). However, current static graph unlearning studies cannot unlearn dynamic graph elements and exhibit limitations such as the model-specific design or reliance on pre-processing, which disenable their practicability in dynamic graph unlearning. To this end, we study the dynamic graph unlearning for the first time and propose an effective, efficient, general, and post-processing method to implement DGNN unlearning. Specifically, we first formulate dynamic graph unlearning in the context of continuous-time dynamic graphs, and then propose a method called Gradient Transformation that directly maps the unlearning request to the desired parameter update. Comprehensive evaluations on six real-world datasets and state-of-the-art DGNN backbones demonstrate its effectiveness (e.g., limited drop or obvious improvement in utility) and efficiency (e.g., 7.23× speed-up) advantages. Additionally, our method has the potential to handle future unlearning requests with significant performance gains (e.g., 32.59× speed-up).

The Web of Things (WoT) enhances interoperability across webbased and ubiquitous computing platforms while complementing existing IoT standards.The multimodal Federated Learning (FL) paradigm has been introduced to enhance WoT by enabling the fusion of multi-source mobile sensing data while preserving privacy.However, a key challenge in mobile sensing systems using multimodal FL is modality incompleteness, where some modalities may be unavailable or only partially captured, potentially degrading the system's performance and reliability.Current multimodal FL frameworks typically train multiple unimodal FL subsystems or apply interpolation techniques on the node side to approximate missing modalities.However, these approaches overlook the shared latent feature space among incomplete modalities across different nodes and fail to discriminate against low-quality nodes.To address this gap, we present FedMobile, a new knowledge contribution-aware multimodal FL framework designed for robust learning despite missing modalities.FedMobile prioritizes local-to-global knowledge transfer, leveraging cross-node multimodal feature information to reconstruct missing features.It also enhances system performance and resilience to modality heterogeneity through rigorous node contribution assessments and knowledge contribution-aware aggregation rules.Empirical evaluations on five widely recognized multimodal benchmark datasets demonstrate that FedMobile maintains robust learning even when up to 90% of modality information is missing or when data from two modalities are randomly missing, outperforming state-of-the-art baselines.Our code and data are available at the link.

Deceptive patterns (DPs) are user interface designs deliberately crafted to manipulate users into unintended decisions, often by exploiting cognitive biases for the benefit of companies or services. While numerous studies have explored ways to identify these deceptive patterns, many existing solutions require significant human intervention and struggle to keep pace with the evolving nature of deceptive designs. To address these challenges, we expanded the deceptive pattern taxonomy from security and privacy perspectives, refining its categories and scope. We created a comprehensive dataset of deceptive patterns by integrating existing small-scale datasets with new samples, resulting in 6,725 images and 10,421 DP instances from mobile apps and websites. We then developed DPGuard, a novel automatic tool leveraging commercial multimodal large language models (MLLMs) for deceptive pattern detection. Experimental results show that DPGuard outperforms state-of-the-art methods. An extensive empirical evaluation on 2,000 popular mobile apps and websites reveals that 25.7% of mobile apps and 49.0% websites feature at least one deceptive pattern instance. Through 4 unexplored case studies that inform security implications, we highlight the critical importance of the unified taxonomy in addressing the growing challenges of Internet deception.

Recent developments in the field of Dynamic Searchable Symmetric Encryption (DSSE) with forward and backward privacy have attracted much attention from both research and industrial communities. However, most DSSE schemes with forward and backward privacy schemes only support single keyword queries, which impedes its prevalence in practice. Although some forward and backward private DSSE schemes with expressive queries (e.g., conjunctive queries) have been introduced, their backward privacy either essentially corresponds to single keyword queries or forward privacy is not comprehensive. In addition, the deletion of many DSSE schemes is achieved by addition paired with a deletion mark (i.e., lazy deletion). To address these problems, we present two novel DSSE schemes with conjunctive queries (termed SDSSE-CQ and SDSSE-CQ-S), which achieve both forward and backward privacy. To analyze their security, we present two new levels of backward privacy (named Type-O and Type-O-, more and more secure), which give a more comprehensive understanding of the leakages of conjunctive queries in the OXT framework. Eventually, the security analysis and experimental evaluations show that the proposed schemes achieve better security with reasonable computation and communication increase.

Organizations these days capitalize on crowdsourcing to learn collective wisdom from a population of individuals. Vast amounts of data have been gathered, making the crowdsourcing platforms a lucrative target to steal data from and thus raising severe privacy concerns. Data contributed by workers may carry sensitive individual information. Meanwhile, organizations deem the aggregate statistics as intellectual property. In this paper, we propose, design, and evaluate GoCrowd, a system framework for obliviously aggregating wisdom with quality assurance in crowdsourcing. At its core, we propose constructions for two procedures. The starting point is a gold-standard based private worker quality control procedure that provides privacy-friendly worker quality assurance under the widely popular gold-standard mechanism. The subsequent procedure is an oblivious wisdom aggregation procedure that obliviously learns aggregate statistics over workers’ data while considering their quality. We securely realize these procedures with only lightweight secret sharing techniques. Our system is utterly oblivious to the service provider, and ensures that only the requester can learn the aggregate quality-aware statistics but nothing more. Extensive evaluations show that GoCrowd can produce quality statistics over data from 500 workers for 200 16-choice questions within 1 s.

With large AI systems and models (LAMs) playing an ever-growing role across diverse applications, their impact on the privacy and cybersecurity of critical infrastructure has become a pressing concern. The LAMPS workshop is dedicated to tackling these emerging challenges, promoting dialogue on cutting-edge developments and ethical issues in safeguarding LAMs within critical infrastructure contexts. Bringing together leading experts from around the world, this workshop will delve into the complex privacy and cybersecurity risks posed by LAMs in critical sectors. Attendees will explore innovative solutions, exchange best practices, and contribute to shaping the future research agenda, emphasizing the crucial balance between advancing AI technologies and securing critical digital and physical infrastructures.

Graph neural networks (GNNs) have gained significant attraction due to their expansive real-world applications. To build trustworthy GNNs, two aspects - fairness and privacy - have emerged as critical considerations. Previous studies have separately examined the fairness and privacy aspects of GNNs, revealing their tradeoff with GNN performance. Yet, the inter-play between these two aspects remains unexplored. In this paper, we pioneer the exploration of the interaction between the privacy risks of edge leakage and the individual fairness of a GNN. Our theoretical analysis unravels that edge privacy risks unfortunately escalate when the nodes' individual fairness improves. Such an issue hinders the accomplishment of privacy and fairness of GNNs at the same time. To balance fairness and privacy, we carefully introduce fairness-aware loss reweighting based on in-fluence function and privacy-aware graph structure perturbation modules within a fine-tuning mechanism. Experimental results underscore the effectiveness of our approach in achieving GNN fairness with limited performance compromise and controlled privacy risks. This work contributes to the comprehensively developing trustworthy GNNs by simultaneously addressing both fairness and privacy aspects.

Federated Learning (FL) is susceptible to poisoning attacks, wherein compromised clients manipulate the global model by modifying local datasets or sending manipulated model updates. Experienced defenders can readily detect and mitigate the poisoning effects of malicious behaviors using Byzantine-robust aggregation rules. However, the exploration of poisoning attacks in scenarios where such behaviors are absent remains largely unexplored for Byzantine-robust FL. This paper addresses the challenging problem of poisoning Byzantine-robust FL by introducing catastrophic forgetting. To fill this gap, we first formally define generalization error and establish its connection to catastrophic forgetting, paving the way for the development of a clean-label data poisoning attack named BadSampler. This attack leverages only clean-label data (i.e., without poisoned data) to poison Byzantine-robust FL and requires the adversary to selectively sample training data with high loss to feed model training and maximize the model's generalization error. We formulate the attack as an optimization problem and present two elegant adversarial sampling strategies, Top-$\kappa$ sampling, and meta-sampling, to approximately solve it. Additionally, our formal error upper bound and time complexity analysis demonstrate that our design can preserve attack utility with high efficiency. Extensive evaluations on two real-world datasets illustrate the effectiveness and performance of our proposed attacks.

Large Vision Language Models (VLMs) extend and enhance the perceptual abilities of Large Language Models (LLMs).Despite offering new possibilities for LLM applications, these advancements raise significant security and ethical concerns, particularly regarding the generation of harmful content.While LLMs have undergone extensive security evaluations with the aid of red teaming frameworks, VLMs currently lack a well-developed one.To fill this gap, we introduce Arondight, a standardized red team framework tailored specifically for VLMs.Arondight is dedicated to resolving issues related to the absence of visual modality and inadequate diversity encountered when transitioning existing red teaming methodologies from LLMs to VLMs.Our framework features an automated multi-modal jailbreak attack, wherein visual jailbreak prompts are produced by a red team VLM, and textual prompts are generated by a red team LLM guided by a reinforcement learning agent.To enhance the comprehensiveness of VLM security evaluation, we integrate entropy bonuses and novelty reward metrics.These elements incentivize the RL agent to guide the red team LLM in creating a wider array of diverse and previously unseen test cases.Our evaluation of ten cutting-edge VLMs exposes significant security vulnerabilities, particularly in generating toxic images and aligning multi-modal prompts.In particular, our Arondight achieves an average attack success rate of 84.5% on GPT-4 in all fourteen prohibited scenarios defined by OpenAI in terms of generating toxic text.For a clearer comparison, we also categorize existing VLMs based on their safety levels and provide corresponding reinforcement recommendations.Our multimodal prompt dataset and red team code will be released after ethics committee approval.

The deployment of Graph Neural Networks (GNNs) within Machine Learning as a Service (MLaaS) has opened up new attack surfaces and an escalation in security concerns regarding model-centric attacks. These attacks can directly manipulate the GNN model parameters during serving, causing incorrect predictions and posing substantial threats to essential GNN applications. Traditional integrity verification methods falter in this context due to the limitations imposed by MLaaS and the distinct characteristics of GNN models.In this research, we introduce a groundbreaking approach to protect GNN models in MLaaS from model-centric attacks. Our approach includes a comprehensive verification schema for GNN’s integrity, taking into account both transductive and inductive GNNs, and accommodating varying pre-deployment knowledge of the models. We propose a query-based verification technique, fortified with innovative node fingerprint generation algorithms. To deal with advanced attackers who know our mechanisms in advance, we introduce randomized fingerprint nodes within our design. The experimental evaluation demonstrates that our method can detect five representative adversarial model-centric attacks, displaying 2 to 4 times greater efficiency compared to baselines.

Graph neural networks (GNNs) have emerged as a series of competent graph learning methods for diverse real-world scenarios, ranging from daily applications such as recommendation systems and question answering to cutting-edge technologies such as drug discovery in life sciences and n-body simulation in astrophysics. However, task performance is not the only requirement for GNNs. Performance-oriented GNNs have exhibited potential adverse effects, such as vulnerability to adversarial attacks, unexplainable discrimination against disadvantaged groups, or excessive resource consumption in edge computing environments. To avoid these unintentional harms, it is necessary to build competent GNNs characterized by trustworthiness. To this end, we propose a comprehensive roadmap to build trustworthy GNNs from the view of the various computing technologies involved. In this survey, we introduce basic concepts and comprehensively summarize existing efforts for trustworthy GNNs from six aspects, including robustness, explainability, privacy, fairness, accountability, and environmental well-being. In addition, we highlight the intricate cross-aspect relations between the above six aspects of trustworthy GNNs. Finally, we present a thorough overview of trending directions for facilitating the research and industrialization of trustworthy GNNs.

Multi-source multi-client (M/M) searchable encryption has drawn increasing attention as data sharing becomes prevalent in the digital economics era. It allows data from multiple sources to be securely outsourced to third parties and queried by authorized clients. In response to these demands, various schemes sprung up in the last few years. However, empirical results show that they suffer from performance limitations. Specifically, they either require per-interaction in per-query between data sources and clients or time-consuming public-key encryption. To address these issues, we propose a searchable encryption scheme that allows authorized clients to efficiently search encrypted data from multiple sources. Compared to previous schemes, our design reduces the interaction overhead of authorization and query with the aid of a set-constrained pseudo-random function. Given practical considerations in the M/M setting, we further refine the forward privacy (FP) as “FP with client” and “FP with server” for data addition. To achieve these new security notions, we construct a new M/M scheme only with efficient symmetric cryptographic tools. We perform a formal security analysis of the proposed schemes and implement them to compare with prior arts. The theoretical and experimental results confirm that our designs are practical with lower communication and computation overhead.

The rapid growth of Decentralized Finance (DeFi) boosts the blockchain ecosystem. At the same time, attacks on DeFi applications (apps) are increasing. However, to the best of our knowledge, existing smart contract vulnerability detection tools cannot directly detect DeFi attacks. That's because they lack the capability to recover and understand high-level DeFi semantics, e.g., a user trades a token pairXandYin a Decentralized EXchange (DEX). In this work, we focus on the detection of two new types of price manipulation attacks. To this end, we propose a platform-independent method to identify high-level DeFi semantics. Specifically, we first construct the Cash Flow Tree (CFT) from a raw transaction and then lifting the low-level semantics to high-level ones, including five advanced DeFi actions. Finally, we use patterns expressed with the recovered DeFi semantics to detect price manipulation attacks. We implemented a prototype namedDeFiRangerthat detected 14zero-daysecurity incidents. These findings were reported to affected parties or/and the community for the first time. Furthermore, the backtest experiment discovered 15 unknown historical security incidents. We further performed an attack analysis to shed light on the root causes of vulnerabilities incurring price manipulation attacks.

The prosperity of Decentralized Finance (DeFi) unveils underlying risks, with reported losses surpassing 3.2 billion USD between 2018 and 2022 due to vulnerabilities in Decentralized Applications (DApps). One significant threat is the Price Manipulation Attack (PMA) that alters asset prices during transaction execution. As a result, PMA accounts for over 50 million USD in losses. To address the urgent need for efficient PMA detection, this article introduces a novel detection service,DeFiGuard, using Graph Neural Networks (GNNs). In this article, we propose cash flow graphs with four distinct features, which capture the trading behaviors from transactions. Moreover,DeFiGuardintegrates transaction parsing, graph construction, model training, and PMA detection. Evaluations on the collected transactions demonstrate thatDeFiGuardwith GNN models outperforms the baseline MLP model and classical classification models in Accuracy, TPR, FPR, and AUC-ROC. The results of ablation studies suggest that the combination of the four proposed node features enhancesDeFiGuard’s efficacy. Moreover,DeFiGuardclassifies transactions within 0.892 to 5.317 seconds, which provides sufficient time for the victims (DApps and users) to take action to rescue their vulnerable funds. In conclusion, this research offers a significant step towards safeguarding the DeFi landscape from PMAs using GNNs.

Dynamic searchable symmetric encryption (DSSE) enables a server to efficiently search and update over encrypted files. To minimize the leakage during updates, a security notion named forward and backward privacy is expected for newly proposed DSSE schemes. Those schemes are generally constructed in a way to break the linkability across search and update queries to a given keyword. However, it remains underexplored whether forward and backward private DSSE is resilient against practical leakage-abuse attacks (LAAs), where an attacker attempts to recover query keywords from the leakage passively collected during queries.

While graph neural networks (GNNs) dominate the state-of-the-art for exploring graphs in real-world applications, they have been shown to be vulnerable to a growing number of privacy attacks. For instance, link stealing is a well-known membership inference attack (MIA) on edges that infers the presence of an edge in a GNN’s training graph. Recent studies on independent and identically distributed data (e.g., images) have empirically demonstrated that individuals from different groups suffer from different levels of privacy risks to MIAs, i.e., uneven vulnerability. However, theoretical evidence of such uneven vulnerability is missing. In this paper, we first present theoretical evidence of the uneven vulnerability of GNNs to link stealing attacks, which lays the foundation for demystifying such uneven risks among different groups of edges. We further demonstrate a group-based attack paradigm to expose the practical privacy harm to GNN users derived from the uneven vulnerability of edges. Finally, we empirically validate the existence of obvious uneven vulnerability on nine real-world datasets (e.g., about 25% AUC difference between different groups in the Credit graph). Compared with existing methods, the outperformance of our group-based attack paradigm confirms that customising different strategies for different groups results in more effective privacy attacks.

The rising popularity of ear-wear devices equipped with motion sensors has brought concerns regarding privacy issues due to their powerful sensing capabilities. Previous studies have shown the potential for speech eavesdropping using earphone motion sensors with a sampling frequency of 1000 Hz. However, as the risks of such attacks continue to escalate, mobile operating systems like Android have imposed limitations on the sampling frequency, typically no more than 200 Hz, to avoid such attacks. The lower sampling frequency reduces the amount of collected information within the same timeframe, potentially leading to decreased accuracy. In this paper, we further investigate the effectiveness of utilizing earphone motion sensors for inferring sensitive information at a sampling frequency of 200 Hz while directly using raw data without any data transformation to prevent information loss. We employ a channel attention mechanism to dynamically adjust axis weights to address the varying energy levels across different sensor axes. Meanwhile, we analyze the impact of sampling frequency, environment, and volume on speech recognition performance. Additionally, we explore the extraction of other information from speech signals, such as speaker identity and gender. Our experiments on two datasets demonstrate high recognition accuracy for all three tasks at the 200Hz sampling frequency. We expect our work to raise awareness among manufacturers regarding the privacy issues associated with earphone motion sensors.

Blockchain-based cloud storage is one of the areas which have been developing rapidly in both academia and industry in recent years. As a key feature to improve the usability of such systems, keyword search is yet missing in existing blockchain-based cloud storage systems due to the encryption of uploaded files. To enable this important feature for blockchain-based cloud storage systems while preserving privacy, blockchain-based Symmetric Searchable Encryption schemes have attracted a lot of research interest. In this research, we pinpoint three important requirements for blockchain-based SSE systems to be practical. The first two requirements are fast query response and low on-chain storage cost. They are the key for a blockchain-based SSE system to be usable and feasible. The third requirement is soundness, which guarantees that dishonest behaviours from both users and service providers can be detected and prevented. To the best of our knowledge, none of the existing studies achieves these properties at the same time. To fill in this gap, we present a novel construction of the blockchain-based SSE system. We provide both theoretical analysis and empirical evaluation to show that the system achieves all the desired properties.

Large-scale language models have achieved tremendous success across various natural language processing (NLP) applications. Nevertheless, language models are vulnerable to backdoor attacks, which inject stealthy triggers into models for steering them to undesirable behaviors. Most existing backdoor attacks, such as data poisoning, require further (re)training or fine-tuning language models to learn the intended backdoor patterns. The additional training process however diminishes the stealthiness of the attacks, as training a language model usually requires long optimization time, a massive amount of data, and considerable modifications to the model parameters.

Neural network (NN) inference services enrich many applications, like image classification, object recognition, facial verification, and more. These NN inference services are increasingly becoming an essential offering from cloud computing providers, where end-users’ data are offloaded to the cloud for inference under a customized model. However, current cloud-based inference services operate on clear inputs and NN models, raising paramount privacy concerns. Individual user data may contain private information that should always remain confidential. Meanwhile, the NN model is deemed proprietary to the model owner as model training requires substantial resources. In this article, we present, tailor, and evaluateSonic, a lightweight secure NN inference service delegated in the cloud.Sonicleverages the cloud computing paradigm to fully outsource the secure inference, freeing end devices and model owners from being actively online for assistance.Sonicguards both user input and model privacy along the whole service flow. We design a series of secure and efficient NN layer functions purely using lightweight cryptographic primitives. Extensive evaluations demonstrate thatSonicachieves up to$60\times$bandwidth saving in online inference compared to prior art.

Transfer learning is prevalent as a technique to efficiently generate new models (Student models) based on the knowledge transferred from a pre-trained model (Teacher model). However, Teacher models are often publicly available for sharing and reuse, which inevitably introduces vulnerability to trigger severe attacks against transfer learning systems. In this article, we take a first step towards mitigating one of the most advanced misclassification attacks in transfer learning. We design a distilleddifferentiatorvia activation-based network pruning to enervate the attack transferability while retaining accuracy. We adopt an ensemble structure from variant differentiators to improve the defence robustness. To avoid the bloated ensemble size during inference, we propose a two-phase defence, in which inference from the Student model is first performed to narrow down the candidate differentiators to be assembled, and later only a small, fixed number of them can be chosen to validate clean or reject adversarial inputs effectively. Our comprehensive evaluations on both large and small image recognition tasks confirm that the Student models with our defence of only 5 differentiators are immune to over 90% of the adversarial inputs with an accuracy loss of less than 10%. Our comparison also demonstrates that our design outperforms prior problematic defences.

Federated learning has recently emerged as a paradigm promising the benefits of harnessing rich data from diverse sources to train high quality models, with the salient features that training datasets never leave local devices. Only model updates are locally computed and shared for aggregation to produce a global model. While federated learning greatly alleviates the privacy concerns as opposed to learning with centralized data, sharing model updates still poses privacy risks. In this paper, we present a system design which offers efficient protection of individual model updates throughout the learning procedure, allowing clients to only provide obscured model updates while a cloud server can still perform the aggregation. Our federated learning system first departs from prior works by supporting lightweight encryption and aggregation, and resilience against drop-out clients with no impact on their participation in future rounds. Meanwhile, prior work largely overlooks bandwidth efficiency optimization in the ciphertext domain and the support of security against an actively adversarial cloud server, which we also fully explore in this paper and provide effective and efficient mechanisms. Extensive experiments over several benchmark datasets (MNIST, CIFAR-10, and CelebA) show our system achieves accuracy comparable to the plaintext baseline, with practical performance.

Cloud storage systems have seen a growing number of clients due to the fact that more and more businesses and governments are shifting away from in-house data servers and seeking cost-effective and ease-of-access solutions. However, the security of cloud storage is underestimated in current practice, which resulted in many large-scale data breaches. To change the status quo, this paper presents the design of ShieldDB, an encrypted document database. ShieldDB adapts the searchable encryption technique to preserve the search functionality over encrypted documents without having much impact on its scalability. However, merely realising such a theoretical primitive suffers from real-world threats, where a knowledgeable adversary can exploit the leakage (aka access pattern to the database) to break the claimed protection on data confidentiality. To address this challenge in practical deployment, ShieldDB is designed with tailored padding countermeasures. Unlike prior works, we target a more realistic adversarial model, where the database gets updated continuously, and the adversary can monitor it at an (or multiple) arbitrary time interval(s). ShieldDB’s padding strategies ensure that the access pattern to the database is obfuscated all the time. We present a full-fledged implementation of ShieldDB and conduct intensive evaluations on Azure Cloud.

Graph neural networks (GNNs) offer promising learning methods for graph-related tasks. However, GNNs are at risk of adversarial attacks. Two primary limitations of the current evasion attack methods are highlighted: (1) The currentGradArgmaxignores the “long-term” benefit of the perturbation. It is faced with zero-gradient and invalid benefit estimates in certain situations. (2) In the reinforcement learning-based attack methods, the learned attack strategies might not be transferable when the attack budget changes. To this end, we first formulate the perturbation space and propose an evaluation framework and the projective ranking method. We aim to learn a powerful attack strategy then adapt it as little as possible to generate adversarial samples under dynamic budget settings. In our method, based on mutual information, we rank and assess the attack benefits of each perturbation for an effective attack strategy. By projecting the strategy, our method dramatically minimizes the cost of learning a new attack strategy when the attack budget changes. In the comparative assessment withGradArgmaxandRL-S2V, the results show our method owns high attack performance and effective transferability. The visualization of our method also reveals various attack patterns in the generation of adversarial samples.

Edge computing brings data computation and storage closer to the mobile device to save response time for decision making. After being processed at the edge, commonly, the data will be uploaded to the cloud for further enriched analysis. For privacy concerns, local devices may encrypt the collected data before sending it to the cloud server. However, this treatment increases the servers processing time and makes it hard to pick out the desired data. In this paper, to address this problem, we design an encrypted data validation scheme, which enables the edge to clean the encrypted data to be uploaded. Because edge computing encompasses numerous edge devices from different service providers, we explore the public-key cryptographic mechanism to implement our secure data validation scheme. Considering potential risks from quantum computers, we propose to leverage ideal lattice to realize our protocol, which reaches better performance in both time and storage for edge devices. Extensive evaluation results show that our proposed proposal achieves considerable performance improvements in terms of communication and computation aspects. Particularly, compared to prior work, nearly 123x-238x speed up is achieved in our key derivation procedure and the storage cost of the secret key is reduced from 547MB to 1.7MB.

Machine learning models are shown to face a severe threat from Model Extraction Attacks, where a well-trained private model owned by a service provider can be stolen by an attacker pretending as a client. Unfortunately, prior works focus on the models trained over the Euclidean space, e.g., images and texts, while how to extract a GNN model that contains a graph structure and node features is yet to be explored. In this paper, for the first time, we comprehensively investigate and develop model extraction attacks against GNN models. We first systematically formalise the threat modelling in the context of GNN model extraction and classify the adversarial threats into seven categories by considering different background knowledge of the attacker, e.g., attributes and/or neighbour connections of the nodes obtained by the attacker. Then we present detailed methods which utilise the accessible knowledge in each threat to implement the attacks. By evaluating over three real-world datasets, our attacks are shown to extract duplicated models effectively, i.e., 84% - 89% of the inputs in the target domain have the same output predictions as the victim model.

In Machine Learning, the emergence of the right to be forgotten gave birth to a paradigm named machine unlearning, which enables data holders to proactively erase their data from a trained model. Existing machine unlearning techniques focus on centralized training, where access to all holders’ training data is a must for the server to conduct the unlearning process. It remains largely underexplored about how to achieve unlearning when full access to all training data becomes unavailable. One noteworthy example is Federated Learning (FL), where each participating data holder trains locally, without sharing their training data to the central server. In this paper, we investigate the problem of machine unlearning in FL systems. We start with a formal definition of the unlearning problem in FL and propose a rapid retraining approach to fully erase data samples from a trained FL model. The resulting design allows data holders to jointly conduct the unlearning process efficiently while keeping their training data locally. Our formal convergence and complexity analysis demonstrate that our design can preserve model utility with high efficiency. Extensive evaluations on four real-world datasets illustrate the effectiveness and performance of our proposed realization.

The prosperity of decentralized finance motivates many investors to profit via trading their crypto assets on decentralized applications (DApps for short) of the Ethereum ecosystem. Apart from Ether (the native cryptocurrency of Ethereum), many ERC20 (a widely used token standard on Ethereum) tokens obtain vast market value in the ecosystem. Specifically, the approval mechanism is used to delegate the privilege of spending users’ tokens to DApps. By doing so, the DApps can transfer these tokens to arbitrary receivers on behalf of the users. To increase the usability, unlimited approval is commonly adopted by DApps to reduce the required interaction between them and their users. However, as shown in existing security incidents, this mechanism can be abused to steal users’ tokens.

[J1C2 Presentation Abstract at IEEE SERVICES 2021 for IEEE Transactions on Services Computing DOI 10.1109/TSC.2021.3111208]

In this paper, we propose CryptMed, a system framework that enables medical service providers to offer secure, lightweight, and accurate medical diagnostic service to their customers via an execution of neural network inference in the ciphertext domain. CryptMed ensures the privacy of both parties with cryptographic guarantees. Our technical contributions include: 1) presenting a secret sharing based inference protocol that can well cope with the commonly-used linear and non-linear NN layers; 2) devising optimized secure comparison function that can efficiently support comparison-based activation functions in NN architectures; 3) constructing a suite of secure smooth functions built on precise approximation approaches for accurate medical diagnoses. We evaluate CryptMed on 6 neural network architectures across a wide range of non-linear activation functions over two benchmark and four real-world medical datasets. We comprehensively compare our system with prior art in terms of end-to-end service workload and prediction accuracy. Our empirical results demonstrate that CryptMed achieves up to respectively 413 ×, 19 ×, and 43 × bandwidth savings for MNIST, CIFAR-10, and medical applications compared with prior art. For the smooth activation based inference, the best choice of our proposed approximations preserve the precision of original functions, with less than 1.2% accuracy loss and could enhance the precision due to the newly introduced activation function family.

Network Function Virtualisation (NFV) advances the adoption of composable software middleboxes. Accordingly, cloud data centres become major NFV vendors for enterprise traffic processing. Due to the privacy concern of traffic redirection to the cloud, secure middlebox systems (e.g., BlindBox) draw much attention; they can process encrypted packets against encrypted rules directly. However, most of the existing systems supporting pattern matching based network functions require the enterprise gateway to tokenise packet payloads via sliding windows. Such tokenisation induces a considerable communication overhead, which can be over 100× to the packet size. To overcome this bottleneck, in this article, we propose the first bandwidth-efficient encrypted pattern matching protocol for secure middleboxes. We resort to a primitive called symmetric hidden vector encryption (SHVE), and propose a variant of it, aka SHVE+, to achieve constant and moderate communication cost. To speed up, we devise encrypted filters to reduce the number of accesses to SHVE+ during matching highly. We formalise the security of our proposed protocol and conduct comprehensive evaluations over real-world rulesets and traffic dumps. The results show that our design can inspect a packet over 20 k rules within 100$\mu$s. Compared to prior work, it brings a saving of 94 percent in bandwidth consumption.

In this article, we introduce a new mechanism for constructing multi-client searchable encryption (SE). By tactfully leveraging the RSA-function, we propose the first multi-client SE protocol that successfully avoids per-query interaction between data owner and client. Therefore, our approach significantly reduces the communication cost by eliminating the need for data owner to authorize client queries at all times. To be compatible with the RSA-based approach, we also present a deterministic and memory-efficient ‘keyword to prime’ hash function, which may be of independent interest. Further, to improve efficiency, we put forward a more generic construction from set-constrained PRFs. The construction not only inherits the merits of our first protocol, but also achieves an enhanced security (against untrusted clients), where colluding attack among clients is also taken into account. Both protocols are instantiated via the recent representative SE protocol by Cashet al.with the support of boolean queries. At last, we implement our proposed protocols and comprehensively evaluate their performance to demonstrate their practicability and scalability.

The advances in machine learning have revealed its great potential for emerging mobile applications such as face recognition and voice assistant. Models trained via a Neural Network (NN) can offer accurate and efficient inference services for mobile users. Unfortunately, the current deployment of such service encounters privacy concerns. Directly offloading the model to the mobile device violates model privacy of the model owner, while feeding user input to the service compromises user privacy. To address this issue, we propose Leia, a lightweight cryptographic NN inference system at the edge. Leia is designed from two mobile-friendly perspectives. First, it leverages the paradigm of edge computing wherein the inference procedure keeps the model closer to the mobile user to foster low latency service. Specifically, Leia’s architecture consists of two non-colluding edge services to obliviously perform NN inference on the encoded user data and model. Second, Leia’s realization makes the judicious use of potentially constrained computational and communication resources in edge devices. We adapt the Binarized Neural Network (BNN), a trending flavor of NN with low inference overhead, and purely choose the lightweight secret sharing techniques to realize secure blocks of BNN. We implement Leia and deploy it on Raspberry Pi. Empirical evaluations on benchmark and medical datasets via various models demonstrate the practicality of Leia.

GPS trajectories serve as a significant data source for travel mode identification along with the development of various GPS-enabled smart devices. However, such data directly integrate user private information, thus hindering users from sharing data with third parties. On the other hand, existing identification methods heavily depend on the respective manual travel mode annotations, whose production is economically inefficient and error-prone. In this paper, we propose a Semi-supervised Federated Learning (SSFL) framework that can accurately identify travel modes without using users’ raw trajectories data or relying on notable data labels. Specifically, we propose a new identification model named convolutional neural network-gated recurrent unit model in SSFL to accurately infer travel modes from GPS trajectories. Second, we design a pseudo-labeling method for the clients to set pseudo-labels on their local unlabeled dataset by using a small public dataset at the server. Furthermore, we adopt a grouping-based aggregation scheme and a data flipping augmentation scheme, which can boost the convergence and performance of the proposed framework. Comprehensive evaluations on a real-world dataset show that SSFL outperforms centralized semi-supervised baselines and is robust to the non-independent and identically distributed data commonly seen in practice.

Graph data has been widely used to represent data from various domain, e.g., social networks, recommendation system. With great power, the GNN models, usually as valuable properties of their owners, also become attractive targets of the adversary who covets to steal them. While existing works show that simple deep neural networks can be reproduced by so-called Model Extraction Attacks, how to extract a GNN model has not been explored. In this paper, we exploit the threat of model extraction attacks against GNN models. Unlike ordinary attacks which obtain model information via only the input-output query pairs, we utilize both the node queries and the graph structure to extract the GNNs. Furthermore, we consider the stealthiness of the attack and propose to generate legitimate queries so the extraction can be applied discreetly. We implement our attack by leveraging the responses of these queries, as well as other accessible knowledge, e.g., neighbor connectives of the queried nodes. By evaluating over three real-world datasets, our attack is shown to effectively produce a surrogate model with more than 80% equivalent predictions as the target model.

Graph Neural Networks (GNNs) have emerged as a series of effective learning methods for graph-related tasks. However, GNNs are shown vulnerable to adversarial attacks, where attackers can fool GNNs into making wrong predictions on adversarial samples with well-designed perturbations. Specifically, we observe that the current evasion attacks suffer from two limitations: (1) the attack strategy based on the reinforcement learning method might not be transferable when the attack budget changes; (2) the greedy mechanism in the vanilla gradient-based method ignores the long-term benefits of each perturbation operation. In this paper, we propose a new attack method named projective ranking to overcome the above limitations. Our idea is to learn a powerful attack strategy considering the long-term benefits of perturbations, then adjust it as little as possible to generate adversarial samples under different budgets. We further employ mutual information to measure the long-term benefits of each perturbation and rank them accordingly, so the learned attack strategy has better attack performance. Our method dramatically reduces the adaptation cost of learning a new attack strategy by projecting the attack strategy when the attack budget changes. Our preliminary evaluation results in synthesized and real-world datasets demonstrate that our method owns powerful attack performance and effective transferability.

In light of the wide application of Graph Neural Networks (GNNs), Membership Inference Attack (MIA) against GNNs raises severe privacy concerns, where training data can be leaked from trained GNN models. However, prior studies focus on inferring the membership of only the components in a graph, e.g., an individual node or edge. In this paper, we take the first step in MIA against GNNs for graph-level classification. Our objective is to infer whether a graph sample has been used for training a GNN model. We present and implement two types of attacks, i.e., training-based attacks and threshold-based attacks from different adversarial capabilities. We perform comprehensive experiments to evaluate our attacks in seven real-world datasets using five representative GNN models. Both our attacks are shown effective and can achieve high performance, i.e., reaching over 0.7 attack F1 scores in most cases1. Our findings also confirm that, unlike the node-level classifier, MIAs on graph-level classification tasks are more co-related with the overfitting level of GNNs rather than the statistic property of their training graphs.

Blockchain has led the trend of decentralized applications and shown great use beyond cryptocurrencies. Decentralized storage such as Storj and Sia leverages blockchain to establish an open platform for sharing economy, which provides private and reliable file-outsourcing services. However, the ubiquitous keyword search function over encrypted files is yet to be supported. To enable this function, we first apply searchable encryption techniques to the decentralized setting. But this primitive can hardly ensure the service integrity. The reason is that decentralized storage commonly faces severe threats from both clients and service peers. Service peers may return partial or incorrect results, while clients may intentionally slander the service peers to avoid payments. To address these threats, we utilize the smart contract to record the logs of encrypted search (aka evidence) on the blockchain, and devise a fair protocol to handle disputes and issue fair payments. Using a dynamic-efficient searchable encryption scheme as an instantiation, we craft a concrete scheme that preserves encrypted search capability and enforces ecosystem healthiness, so that service peers are incentivized to make real efforts and jointly guarantee service reliability. We implement our scheme in Python and Solidity, and test its search performance and transaction costs on Ethereum.

Searchable symmetric encryption (SSE) enables users to make confidential queries over always encrypted data while confining information disclosure to pre-defined leakage profiles. Despite the well-understood performance and potentially broad applications of SSE, recent leakage-abuse attacks (LAAs) are questioning its real-world security implications. They show that a passive adversary with certain prior information of a database can recover queries by exploiting the legitimately admitted leakage. While several countermeasures have been proposed, they are insufficient for either security, i.e., handling only specific leakage like query volume, or efficiency, i.e., incurring large storage and bandwidth overhead. We aim to fill this gap by advancing the understanding of LAAs from a fundamental algebraic perspective. Our investigation starts by revealing that the index matrices of a plaintext database and its encrypted image can be linked by linear transformation. The invariant characteristics preserved under the transformation encompass and surpass the information exploited by previous LAAs. They allow one to unambiguously link encrypted queries with corresponding keywords, even with only partial knowledge of the database. Accordingly, we devise a new powerful attack and conduct a series of experiments to show its effectiveness. In response, we propose a new security notion to thwart LAAs in general, inspired by the principle of local differential privacy (LDP). Under the notion, we further develop a practical countermeasure with tunable privacy and efficiency guarantee. Experiment results on representative real-world datasets show that our countermeasure can reduce the query recovery rate of LAAs, including our own.

In-network storage is recognized as a vital component of many emerging network architectures, which facilitates high-quality and efficient content-centric services. In this trend, providing content-based near-duplicate detection (NDD) services among in-network storage becomes naturally necessary for network traffic alleviation and resource optimization. However, due to the increasing attacking surfaces, storing data in the networked environment inevitably raises new concerns about user privacy exposure and unauthorized data access. Therefore, we aim to design a secure NDD service in the context of encrypted in-network storage. For efficiency, we first leverage the fingerprint techniques and locality-sensitive hashing to convert the problem of NDD into the keyword search. We then adopt an efficient multi-key searchable encryption scheme, which requires only one encrypted query from the user even the data are from multiple content providers encrypted with different keys. As simply combining the above methods does not appear to directly locate accurate results, we then devise a secure result refining scheme via Yao's garbled circuits to avoid user-side post-processing. Furthermore, we enhance our design to address the potential malicious behavior of in-network servers. Extensive evaluations of real-world image dataset demonstrate that our design can achieve comparable accuracy to the plaintext with modest security overhead.

The explosive growth of network traffic is pushing forward the paradigm of cloud-based middlebox services today. However, due to the increasing attacking surfaces, redirecting enterprises traffic to outsourced middleboxes inevitably raises new privacy concerns about packet content exposure and unauthorized rulesets access. To address these issues, recent efforts have been made toward enabling middlebox services through encrypted traffic and middlebox rules. Following this direction, in this article, we investigate the issue of privacy-preserving header checking, which is an indispensable service of middlebox applications. Specifically, we propose two new encrypted header-matching schemes that significantly improve security and efficiency. Our main idea is to formulate the problem of encrypted header checking as range-based pattern matching, and carefully craft security designs to enable efficient header inspection in the ciphertext domain. Our first design is carefully tailored to generic range-based functions, while our second design is highly customized for contiguous rulesets to further improve the checking efficiency. We formally analyze the security strengths and implement a fully functional system prototype. The extensive experiments over the real-world rulesets demonstrate the practicality of our designs.

Outlier detection is widely used in practice to track the anomaly on incremental data sets, such as network traffic and system logs. However, these data sets often involve sensitive information, and sharing the data to third parties for anomaly detection raises privacy concerns. In this article, we present a privacy-preserving outlier detection (PPOD) protocol for incremental data sets. The protocol decomposes the outlier detection algorithm into several phases and recognizes the necessary cryptographic operations in each phase. It realizes several cryptographic modules via efficient and interchangeable protocols to support the above cryptographic operations and composes them in the overall protocol to enable outlier detection over encrypted data sets. To support efficient updates, it integrates the sliding window model to periodically evict the expired data in order to maintain a constant update time. We build a prototype of PPOD and systematically evaluates the cryptographic modules and the overall protocols under various parameter settings. Our results show that PPOD can handle encrypted incremental data sets with a moderate computation and communication cost.

Truth discovery in mobile crowdsensing has recently received wide attention. It refers to the procedure for estimating the unknown user reliability from collected sensory data and inferring truthful information via reliability-aware data aggregation. Though widely studied in the plaintext domain, truth discovery remains largely under-explored in privacy-aware mobile crowdsensing. Existing works either do not consider user reliability issue or fall short of achieving practical cost efficiency, due to iterative transmission and computation over large ciphertexts from homomorphic cryptosystem. In this paper, we propose two new privacy-aware crowdsensing designs with truth discovery that significantly improve the bandwidth and computation performance on individual users. Our insight is to identify the core atomic operation in the iterative truth discovery procedure, and carefully craft security designs accordingly to enable efficient truth discovery in the ciphertext domain. Our first design is highly customized for the single-server setting, while our second design under the two-server model further shifts most of user workloads to the cloud server side. Both our designs protect individual sensory data and reliability degrees throughout the truth discovery procedure. Experiments show that compared with the prior result, our designs gain at least 30x and 10x savings on user communication and computation, respectively.

The persistent growth of big data applications has being raising new challenges in managing large volumes of datasets with high scalability, confidentiality protection, and flexible types of search queries. In this paper, we propose a secure design to disassemble the private dataset with the aim to store them across geographically distributed servers while supporting secure multi-client Boolean queries. In this design, the data owner encrypts the private database with the searchable index attributes. The encrypted dataset will be disassembled and distributed evenly across multiple servers by leveraging the property of a distributed index framework. By constructing an encryption structure, generating search tokens, and enabling parallel query, we show how the proposed design performs the secure while efficient Boolean search. These queries are not only limited to those initiated by the data owner but also can be extended to support multiple authorized clients, where each client is allowed to access a necessary part of the private database. In this stage, we advocate a non-interactive authorization scheme where data owner is not required to stay online to process the query request. Moreover, the query operation can be executed in parallel, which significantly improves the search efficiency. We formally characterize the leakage profile, which allow us to follow the existing security analysis method to demonstrate that our system can guarantee data confidentiality and query privacy. To validate our protocol, we implement a system prototype and evaluate the efficiency of our construction. Through experimental results, we demonstrate the effectiveness of our protocol in terms of data outsourcing time and Boolean query time.

With the prosperity of mobile application markets, mobile advertising is becoming an increasingly important economic force. In order to maximize revenue, ads are recommended to be delivered to potentially interested users, which requires user targeting, i.e., analyzing users' profiles and exploring users' interests. However, collecting user personal information for targeted mobile advertising services raises critical privacy concerns. Although some solutions like anonymization and obfuscation have been proposed for privacy-aware targeted advertising, they undesirably face the issues of security, efficiency, and/or ad relevance. In this paper, we propose a practical system enabling secure and efficient targeted mobile advertising services. It allows the ad network to perform accurate user targeting, while ensuring strong privacy protection for mobile users. Specifically, we show how to properly leverage a cryptographic primitive called private stream searching to support secure, accurate, and practical targeted mobile ad delivery. Moreover, we propose secure billing schemes to enable the ad network to charge advertisers in a privacy-preserving manner. The security strength of our system is thoroughly analyzed. Through extensive experiments, we show that our system achieves practical efficiency on mobile devices.

Running off-site software middleboxes at third-party service providers has been a popular practice. However, routing large volumes of raw traffic, which may carry sensitive information, to a remote site for processing raises severe security concerns. Prior solutions often abstract away important factors pertinent to real-world deployment. In particular, they overlook the significance of metadata protection and stateful processing. Unprotected traffic metadata like low-level headers, size and count, can be exploited to learn supposedly encrypted application contents. Meanwhile, tracking the states of 100,000s of flows concurrently is often indispensable in production-level middleboxes deployed at real networks.

In this paper, we propose GraphSE\textsuperscript2, an encrypted graph database for online social network services to address massive data breaches. GraphSE\textsuperscript2 ~preserves the functionality of social search, a key enabler for quality social network services, where social search queries are conducted on a large-scale social graph and meanwhile perform set and computational operations on user-generated contents. To enable efficient privacy-preserving social search, GraphSE\textsuperscript2 ~provides an encrypted structural data model to facilitate parallel and encrypted graph data access. It is also designed to decompose complex social search queries into atomic operations and realise them via interchangeable protocols in a fast and scalable manner. We build GraphSE\textsuperscript2 ~with various queries supported in the Facebook graph search engine and implement a full-fledged prototype. Extensive evaluations on Azure Cloud demonstrate that GraphSE\textsuperscript2 ~is practical for querying a social graph with a million of users.

Multi-Writer Searchable Encryption, also known as public-key encryption with keyword search(PEKS), serves a wide spectrum of data sharing applications. It allows users to search over encrypted data encrypted via different keys. However, most of the existing PEKS schemes are built on classic security assumptions, which are proven to be untenable to overcome the threats of quantum computers. To address the above problem, in this paper, we propose a lattice-based searchable encryption scheme from the learning with errors (LWE) hardness assumption. Specifically, we observe that the keys of each user in a basic scheme are composed of large-sized matrices and basis of the lattice. To reduce the complexity of key management, our scheme is designed to enable users to directly use their identity for data encryption. We present several optimization techniques for implementation to make our design nearly practical. For completeness, we conduct rigorous security, complexity, and parameter analysis on our scheme, and perform comprehensive evaluations at a commodity machine. With a scenario of 100 users, the cost of key generation for each user is 125s, and the cost of searching a document with 1000 keywords is 13.4ms.

Data outsourcing to cloud has been a common IT practice nowadays due to its significant benefits. Meanwhile, security and privacy concerns are critical obstacles to hinder the further adoption of cloud. Although data encryption can mitigate the problem, it reduces the functionality of query processing, e.g., disabling SQL queries. Several schemes have been proposed to enable one-dimensional query on encrypted data, but multi-dimensional range query has not been well addressed. In this paper, we propose a secure and scalable scheme that can support multi-dimensional range queries over encrypted data. The proposed scheme has three salient features: (1) Privacy: the server cannot learn the contents of queries and data records during query processing. (2) Efficiency: we utilize hierarchical cubes to encode multi-dimensional data records and construct a secure tree index on top of such encoding to achieve sublinear query time. (3) Verifiability: our scheme allows users to verify the correctness and completeness of the query results to address server's malicious behaviors. We perform formal security analysis and comprehensive experimental evaluations. The results on real datasets demonstrate that our scheme achieves practical performance while guaranteeing data privacy and result integrity.

Searchable encryption (SE) is a practical crypto-graphic primitive to build encrypted databases. Recently there has been much attention in leakage-abuse attacks against SE. Among others, attacks based on inference of keyword frequency can easily identify query keywords from the access pattern, i.e., query results. To mitigate these attacks, database padding is considered as a conceptually simple yet effective counter-measure. Unfortunately, none of the existing studies formally understand the relationship between padding security strength and its overhead. Also, how to craft padding is not restricted in current countermeasures, where bogus files are likely to be distinguishable from real ones. In this paper, we propose an information theory based framework to analyse the security strength under certain padding overhead. First, we leverage relative entropy to measure the “closeness” between the distributions of the original dataset and padded dataset. Second, we quantity the attack efforts against padding countermeasures by entropy analysis. Apart from theoretical findings, we further devise an algorithm via outlier detection for padding generation, which considers both the padded dataset distribution and the similarity between real and bogus files. Evaluations on a real-world dataset confirm our theoretical results and demonstrate the efficiency and effectiveness of our proposed padding generation algorithm.

Fog computing has emerged as a promising paradigm in overcoming the growing challenges (e.g., low latency, location awareness, and geographic distribution) arising from many real-world Internet of Things (IoT) applications, by extending the cloud to the network edge. With the widespread deployment of fog-assisted IoT applications, unprecedentedly huge volumes of network traffic from massive IoT devices would continuously arrive at the fog nodes. Archiving the network traffic can be highly beneficial to fog computing, which forms the basis of forensic, monitoring, troubleshooting, and many other critical tasks. Such high value, however, constantly renders traffic archives the first-order target to experienced attackers. This mandates the traffic archives to be built in a trustworthy way and stayed encrypted at rest. Security aside, it is yet highly desirable to retain the utility of the encrypted traffic archives, in particular by making them privately queryable. In this paper, we take the first research attempt and explore a new design point to delicately bridge trusted hardware and searchable encryption for building trustworthy, encrypted, yet queryable network traffic archives for fog-assisted IoT applications. We take a systematic approach to address several key challenges, which are unsolvable by synthesizing out-of-box techniques, from ground up. Extensive evaluations show that our system can achieve stable archiving throughput of 350 Mb/s with one core, and saturate a 1 Gb/s link with four cores; for a real trace, it outperforms a baseline system without any of our designs by over 110× .

The ever-growing of Internet of Things (IoT) data and the new spectrum of data applications have stimulated IoT clients to outsource their data to cloud servers or datacenters. Apart from storage service, the IoT clients also desires the servers to execute functional operations per client's request. In this paper, we aim to design the secure mechanisms that allow the IoT clients to outsource their encrypted data to geographically distributed servers while supporting homomorphic computation functions. We leverage the distributed index framework to disassemble and spread data evenly across geographically distributed servers while employing the key-value store as the underlying structure for fast data retrieval. To support computing over encrypted data, we customize Shamir's secret sharing into our mechanisms to design a tunable scheme for the adaption of different IoT application scenarios. In particular, we design three tunable protocols to achieve the effective additive homomorphic computations while approaching efficiency in terms of servers utilization, computation, and storage overhead. Even the designs focus on the additive computation, we show that it can be readily extended to other types of homomorphic computations as well as verifying the correctness of stored data. Based on the proposed protocols, we design system prototypes, deploy them in Amazon Web services, and evaluate our construction experimentally. Through experimental results, we show that our designs can achieve the efficiency in various perspectives.

To accommodate massive digital data, distributed data stores have become the main solution for cloud services. Among others, key-value stores are widely adopted due to their superior performance. But with the rapid growth of cloud storage, there are growing concerns about data privacy. In this paper, we design and build EncKV, an encrypted and distributed key-value store with rich query support. First, EncKV partitions data records with secondary attributes into a set of encrypted key-value pairs to hide relations between data values. Second, EncKV uses the latest cryptographic techniques for searching on encrypted data, i.e., searchable symmetric encryption (SSE) and order-revealing encryption (ORE) to support secure exact-match and range-match queries, respectively. It further employs a framework for encrypted and distributed indexes supporting query processing in parallel. To address inference attacks on ORE, EncKV is equipped with an enhanced ORE scheme with reduced leakage. For practical considerations, EncKV also enables secure system scaling in a minimally intrusive way. We complete the prototype implementation and deploy it on Amazon Cloud. Experimental results confirm that EncKV preserves the efficiency and scalability of distributed key-value stores.

Symmetric Searchable Encryption (SSE) has received wide attention due to its practical application in searching on encrypted data. Beyond search, data addition and deletion are also supported in dynamic SSE schemes. Unfortunately, these update operations leak some information of updated data. To address this issue, forward-secure SSE is actively explored to protect the relations of newly updated data and previously searched keywords. On the contrary, little work has been done in backward security, which enforces that search should not reveal information of deleted data. In this paper, we propose the first practical and non-interactive backward-secure SSE scheme. In particular, we introduce a new form of symmetric encryption, named symmetric puncturable encryption (SPE), and construct a generic primitive from simple cryptographic tools. Based on this primitive, we then present a backward-secure SSE scheme that can revoke a server's searching ability on deleted data. We instantiate our scheme with a practical puncturable pseudorandom function and implement it on a large dataset. The experimental results demonstrate its efficiency and scalability. Compared to the state-of-the-art, our scheme achieves a speedup of almost 50x in search latency, and a saving of 62% in server storage consumption.

In crowd sensing, truth discovery (TD) refers to finding reliable information from noisy/biased data collected from different providers. To protect providers' data while enabling truth distillation, privacy-preserving truth discovery (PPTD) has received wide attention recently. However, all existing approaches require iterative interaction between server(s) and individual providers, which inevitably demand all providers to be always online. Otherwise, the protocol would fail or expose extra provider information. In this paper, we design and implement the first non-interactive PPTD system that completely removes the online requirement with strong privacy guarantees. Our framework follows the same two-server model from the best-known prior solution, and leverages Yao's Garbled Circuit (GC). Yet, we devise non-trivial speedup techniques for TD-optimized implementation. Firstly, we identify reusable computations in TD to accelerate the circuit generation. Secondly, we securely evaluate the burdensome non-linear functions in TD via customized approximation with accuracy and improved efficiency. Thirdly, we reduce the online execution time by bridging together latest advancements of component-based GC and various computations needed in TD. Unlike prior arts, our framework does not reveal any intermediate results, and further supports “late-join” providers without protocol suspension/restart. The practical performance of our proof-of-concept implementation is verified through extensive evaluations.

Over the past few years, enterprises start adopting software middlebox services from cloud or NFV service providers. Although this new service model is recognized to be cost-effective and scalable for traffic processing, privacy concerns arise because of traffic redirection to outsourced middleboxes. To ease these concerns, recent efforts are made to design secure middlebox services that can directly function over encrypted traffic and middlebox rules. But prior designs only work for portions of frequently-used network functions. To push forward this area, in this work, we investigate header matching based functions like firewall filtering and packet classification. To enable privacy-preserving processing on encrypted packets, we start from the latest primitive “order-revealing encryption (ORE)” for encrypted range search. In particular, we devise a new practical ORE construction tailored for network functions. The advantages include: 1) guaranteed protection of packet headers and rule specified ranges; 2) reduced accessible information during comparisons; 3) rule-aware size reduction for ORE ciphertexts. We implement a fully functional system prototype and deploy it at Microsoft Azure Cloud. Evaluation results show that our system can achieve per packet matching latency 0.53 to 15.87 millisecond over 1.6K firewall rules.

Images are becoming one of the key enablers of user connectivity in social media applications. Many of them are directly exploring image content to suggest new friends with similar interests. To handle the explosive volumes of images, one common trend is to leverage the public cloud as their robust service backend. Despite the convenience, exposing content-rich images to the cloud inevitably raises acute privacy concerns. In this paper, we propose a privacy-preserving architecture for image-centric social discovery services, designed to function over encrypted images. We first adopt the effective Bag-of-Words model to extract the “visual content” of users' images into respective image profile vectors. We then model the core problem as similarity retrieval of encrypted high-dimensional vectors. To achieve scalable services over millions of encrypted images, we design a secure and efficient index structure, which enables practical and accurate social discovery from the cloud, without revealing any image profile or image content. For completeness, we further enrich our service with secure updates, facilitating user's image update. Our implementation is deployed at an Android phone and Amazon Cloud, and extensive experiments are conducted on a large Flickr image dataset which demonstrates the desired quality of services.

In order to ease the burden of the in-house integrated circuit (IC) design, cloud-based IC design platforms advance rapidly, bringing benefits such as reduced capital costs and convenient design collaboration. However, such migration raises security challenges on IC Intellectual Property (IP) protection. Sensitive design data is unwillingly exposed to the cloud. In this paper, we initiate the first study for secure cloud-based IC design, and propose a hybrid cloud framework for privacy-assured IC timing analysis, i.e., an expensive procedure in the IC design flow for circuit delay evaluation. Our key observation is that more and more IP blocks are universally reused. After carefully extracting a small portion of sensitive blocks from the circuit, our framework only outsources non-sensitive design data to the public cloud. However, that “data splitting” hinders sequential delay evaluation. We then develop algorithms to enable the public cloud to derive intermediate results from non-sensitive data, which can be integrated with sensitive data at the private cloud. Additionally, we devise a practical verification protocol to assure the integrity of outsourced computation. Security analysis shows that our design is resilient to IC reverse engineering. Evaluations over large IC benchmarks demonstrate its efficiency and effectiveness.

Searchable Symmetric Encryption (SSE) has been widely studied in cloud storage, which allows cloud services to directly search over encrypted data. Most SSE schemes only work with honest-but-curious cloud services that do not deviate from the prescribed protocols. However, this assumption does not always hold in practice due to the untrusted nature in storage outsourcing. To alleviate the issue, there have been studies on Verifiable Searchable Symmetric Encryption (VSSE), which functions against malicious cloud services by enabling results verification. But to our best knowledge, existing VSSE schemes exhibit very limited applicability, such as only supporting static database, demanding specific SSE constructions, or only working in the single-user model. In this paper, we propose GSSE, the first generic verifiable SSE scheme in the single-owner multiple-user model, which provides verifiability for any SSE schemes and further supports data updates. To generically support result verification, we first decouple the proof index in GSSE from SSE. We then leverage Merkle Patricia Tree (MPT) and Incremental Hash to build the proof index with data update support. We also develop a timestamp-chain for data freshness maintenance across multiple users. Rigorous analysis and experimental evaluations show that GSSE is secure and introduces small overhead for result verification.

Distributed data stores have been rapidly evolving to serve the needs of large-scale applications such as online gaming and real-time targeting. In particular, distributed key-value stores have been widely adopted due to their superior performance. However, these systems do not guarantee to provide strong protection of data confidentiality, and as a result fall short of addressing serious privacy concerns raised from massive data breaches.

Emerging decentralized storage services such as Storj and Filecoin show promise as a new paradigm for data outsourcing. These services tie cryptocurrency to personal storage resources and leverage blockchain technology to ensure data integrity in distributed networks. Compared to current cloud storage, they are expected to be more scalable, cost effective, and secure. In addition to the features above, strong guarantees of data privacy are seriously desired due to today's prevalent data leak and abuse incidents. However, simply using end-to-end encryption limits the search capability and thus will degrade the user experience. In this paper, we propose an encrypted decentralized storage architecture that can support trustworthy and private keyword search functions. We start from searchable encryption to achieve search on encrypted data. Yet, only adopting this primitive is not sufficient to address particular threats in our target decentralized service model. Service peers would maliciously return incorrect results, while user peers would fraudulently refuse to pay service fees. To resolve those threats, we devise specific secure data addition and keyword search protocols to enable client-side verifiability and blockchain based fair judgments on the search results. For practical considerations, we integrate an efficient dynamic searchable encryption scheme to our protocols as an instantiation to lower the blockchain overhead. Our security and performance analysis indicates the advance of the proposed architecture.

Locality sensitive hashing (LSH) is an efficient method for solving the problem of approximate similarity search in high-dimensional spaces. Through LSH, a high-dimensional similarity join can be processed in the same way as hash join, making the cost of joining two large datasets linear. By judicially analyzing the properties of multiple LSH algorithms, we propose a generic method to accelerate the process of joining two large datasets using LSH. The crux of our method lies in the way we identify a set of representative points to reduce the number of LSH lookups. Theoretical analyses show that our proposed method can greatly reduce the number of lookup operations and retain the same result accuracy compared to executing LSH lookups for every query point. Furthermore, we demonstrate the generality of our method by showing that the same principle can be applied to LSH algorithms for three different metrics: the Euclidean distance (QALSH), Jaccard similarity measure (MinHash), and Hamming distance (sequence hashing). Results from experimental studies using real datasets confirm our error analyses and show significant improvements of our method over the state-of-the-art LSH method: to achieve over 0.95 recall, we only need to operate LSH lookups for at most 15% of the query points.

Similarity-oriented services serve as a foundation in a wide range of data analytic applications such as machine learning, target advertising, and real-time decisions. Both industry and academia strive for efficient and scalable similarity discovery and querying techniques to handle massive, complex data records in the real world. In addition to performance, data security and privacy become an indispensable criterion in the quality of service due to progressively increased data breaches. To address this serious concern, in this paper, we propose and implement “EncSIM”, an encrypted and scalable similarity search service. The architecture of EncSIM enables parallel query processing over distributed, encrypted data records. To reduce client overhead, EncSIM resorts to a variant of the state-of-the-art similarity search algorithm, called all-pairs locality-sensitive hashing (LSH). We describe a novel encrypted index construction for EncSIM based on searchable encryption to guarantee the security of service while preserving performance benefits of all-pairs LSH. Moreover, EncSIM supports data record addition with a strong security notion. Intensive evaluations on a cluster of Redis demonstrate low client cost, linear scalability, and satisfied query performance of EncSIM.

Similarity search on high-dimensional data has been intensively studied for data processing and analytics. Despite its broad applicability, data security and privacy concerns along the trend of data outsourcing have not been fully addressed. In this paper, we investigate privacy-preserving similarity join queries, i.e., a pivotal primitive of similarity search that finds pairwise similar data points across two data sets. We start from locality-sensitive hashing and searchable symmetric encryption, i.e., the most practical techniques for similarity search and encrypted search, respectively. However, the immediate combination of two techniques discloses the distribution of the query set, which is exploitable to compromise the confidentiality of queries. To enhance the security, we propose the frequency hiding query scheme, which allows the server to see the flattened query distribution only. To improve the scalability, we further design the result sharing query scheme, which processes a small portion of query points and shares the results with other nearby points. Besides, we set up a strict constraint to carefully select query points to achieve “as-strong-as-possible” guarantees. We formalize the leakage functions in the context of similarity joins, and conduct rigorous security analysis. We implement and evaluate the proposed query schemes on Azure cloud. Experimental results indicate that they have different tradeoffs on security, efficiency, and accuracy, which can flexibly be used for different deployment scenarios.

Locality sensitive hashing (LSH) is an efficient method for solving the problem of approximate similarity search in highdimensional spaces. Through LSH, a high-dimensional similarity join can be processed in the same way as hash join, making the cost of joining two large datasets linear. By judicially analyzing the properties of multiple LSH algorithms, we propose a generic method to speed up the process of joining two large datasets using LSH. The crux of our method lies in the waywhich we identify a set of representative points to reduce the number of LSH lookups. Theoretical analyzes show that our proposed method can greatly reduce the number of lookup operations and retain the same result accuracy compared to executing LSH lookups for every query point. Furthermore, we demonstrate the generality of our method by showing that the same principle can be applied to LSH algorithms for three different metrics: the Euclidean distance (QALSH), Jaccard similarity measure (MinHash), and Hamming distance (sequence hashing). Results from experimental studies using real datasets confirm our error analyzes and show significant improvements of our method overthe state-of-the-art LSH method: to achieve over 0.95 recall, we only need to operate LSH lookups for at most 15 percent of the query points.

Nowadays, large volumes of multimedia data are outsourced to the cloud to better serve mobile applications. Along with this trend, highly correlated datasets can occur commonly, where the rich information buried in correlated data is useful for many cloud data generation/dissemination services. In light of this, we propose to enable a secure and efficient cloud-assisted image sharing architecture for mobile devices, by leveraging outsourced encrypted image datasets with privacy assurance. Different from traditional image sharing, we aim to provide a mobile-friendly design that saves the transmission cost for mobile clients, by directly utilizing outsourced correlated images to reproduce the image of interest inside the cloud for immediate dissemination. First, we propose a secure and efficient index design that allows the mobile client to securely find from encrypted image datasets the candidate selection pertaining to the image of interest for sharing. We then design two specialized encryption mechanisms that support secure image reproduction from encrypted candidate selection. We formally analyze the security strength of the design. Our experiments explicitly show that both the bandwidth and energy consumptions at the mobile client can be saved, while achieving all service requirements and security guarantees.

The explosive growth of multimedia contents, especially videos, is pushing forward the paradigm of cloud-based media hosting today. However, the wide attacking surface of the public cloud and the growing security awareness from the society are both calling for data encryption before outsourcing to cloud. Under the circumstance of encrypted videos, how to still preserve all the service benefits of cloud media center remains to be fully explored. In this paper, we present a secure system architecture design as our initial effort toward this direction, which bridges together the advancements of video coding techniques and secure deduplication. Our design enables the cloud with the crucial deduplication functionality to completely eliminate the extra storage and bandwidth cost, which would have been incurred by hosting encrypted videos from different entities. The design is also carefully tailored to the scalable video coding (SVC) techniques to support heterogeneous networks and devices for high-quality adaptive video dissemination. We show fully functional system implementations with structure-aware encryption design and structure-aware deduplication strategies that are both completely compliant with the video format in SVC. Extensive security analysis and experiments via our prototype deployed on Azure cloud platform show the practicality of the design. Our work can also be easily extended to support other media applications that employ media files with scalable structures.

Modern distributed key-value stores are offering superior performance, incremental scalability, and fine availability for data-intensive computing and cloud-based applications. Among those distributed data stores, the designs that ensure the confidentiality of sensitive data, however, have not been fully explored yet. In this paper, we focus on designing and implementing an encrypted, distributed, and searchable key-value store. It achieves strong protection on data privacy while preserving all the above prominent features of plaintext systems. We first design a secure data partition algorithm that distributes encrypted data evenly across a cluster of nodes. Based on this algorithm, we propose a secure transformation layer that supports multiple data models in a privacy-preserving way, and implement two basic APIs for the proposed encrypted key-value store. To enable secure search queries for secondary attributes of data, we leverage searchable symmetric encryption to design the encrypted secondary indexes which consider security, efficiency, and data locality simultaneously, and further enable secure query processing in parallel. For completeness, we present formal security analysis to demonstrate the strong security strength of the proposed designs. We implement the system prototype and deploy it to a cluster at Microsoft Azure. Comprehensive performance evaluation is conducted in terms of Put/Get throughput, Put/Get latency under different workloads, system scaling cost, and secure query performance. The comparison with Redis shows that our prototype can function in a practical manner.

Migrating middleboxes to third-party service providers (e.g., clouds and ISPs) has drawn widespread attentions recently from both industry and academia. While its benefits on reduced local cost and increased service scalability are well understood, such deployment also introduces new security concerns, due to the fact that these boxes are no longer under the direct control of enterprises. Among others, one fundamental desideratum here is to ensure that those middleboxes consistently perform network functions as intended. In this work, we propose practical solutions towards enabling runtime execution assurances of outsourced middleboxes with high confidence. As an initial effort, we target on pattern matching based network functions, which cover a broad class of middlebox applications such as instruction detection, web firewall, and traffic classification. For efficiency, our design follows the same roadmap of probabilistic checking that provides tunable levels of assurance, as in outsourced computation and distributed computing literature. We show how to synthesize the design intuitions in the context of outsourced middleboxes and the dynamic network effect. We present diligent technical instantiations, in the case of single middlebox and the composition of multiple middlebox service chaining, respectively. For a large batch of packets, sufficiently high assurance levels can be achieved by pre-processing only a few randomly selected packets, with marginal overhead. Evaluations of our system prototype on Amazon EC2 show that, the processing of 1000 packets, which includes pattern matching and execution proof generation, results in 200–500ms latency and throughput up to 360Mbps.

Outsourced middlebox services have drawn broad attentions recently from both industry and academia [1]. Despite benefiting enterprises from reduced cost and increased service scalability, such services also introduce acute security concerns, because these boxes are no longer under direct control of enterprises. Among others, one fundamental and immediate requirement is to ensure that those middleboxes always perform network functions truthfully and correctly [2]. Fulfilling this requirement will extend enterprises' visibility into remote middleboxes and promote further adoption of middlebox outsourcing services. Unfortunately, to our best knowledge, little work investigates the above problem, i.e., making network functions executed by middleboxes verifiable.

Near-duplicate detection (NDD) plays an essential role for effective resource utilization and possible traffic alleviation in many emerging network architectures, leveraging in-network storage for various content-centric services. As innetwork storage grows, data security has become one major concern. Though encryption is viable for in-network data protection, current techniques are still lacking for effectively locating encrypted near-duplicate data, making the benefits of NDD practically invalidated. Besides, adopting encrypted innetwork storage further complicates the user authorization when locating near-duplicate data from multiple content providers under different keys. In this paper, we propose a secure and effective NDD system over encrypted in-network storage supporting multiple content providers. Our design bridges locality-sensitive hashing (LSH) with a newly developed cryptographic primitive, multi-key searchable encryption, which allows the user to send only one encrypted query to access near-duplicate data encrypted under different keys. It relieves the users from multiple rounds of interactions or sending multiple different queries respectively. As simply applying LSH does not ensure the detection quality, we then leverage Yao's garbled circuits to build a secure protocol to obtain highly accurate results, without user-side post-processing. We formally analyze the security strength. Experiments demonstrate our system achieves practical performance with comparable accuracy to plaintext.

Middleboxes are essential for a wide range of advanced traffic processing in modern enterprise networks. Recent trend of deploying middleboxes in cloud as virtualized services further expands potential benefits of middleboxes while avoiding local maintenance burdens. Despite promising, designing outsourced middleboxes still faces several security challenges. First, many middlebox processing services, like intrusion detection, require packet payload inspection, while the ever-increasing adoption of HTTPS limits the function due to the end-to-end encryption. Second, many packet inspection rules used by middleboxes can be proprietary in nature. They may contain sensitive information of enterprises, and thus need strong protection when configuring middleboxes in untrusted outsourced environments. In this paper, we propose a practical system architecture for outsourced middleboxes to perform deep packet inspection over encrypted traffic, without revealing either packet payloads or inspection rules. Our first design is an encrypted high-performance rule filter that takes randomized tokens from packet payloads for encrypted inspection. We then elaborate through carefully tailored techniques how to comprehensively support open-source real rulesets. We formally analyze the security strength. Implementations at Amazon Cloud show that our system introduces roughly 100 millisecond latency in each connection initialization, with individual processing throughput over 3500 packets/second for 500 concurrent connections.

In-network content caching has been a natural trend in emerging network architectures to handle the exponential growth of video traffic. However, due to the potentially wide attacking surfaces, caching video content in the increasingly untrusted networked environment inevitably raises new concerns on user privacy exposure and unauthorized video access. Existing encrypted protocols like HTTPs either fall short of fully leveraging in-network caching or require decrypting the traffic in the middle without guaranteeing the end-to-end security. In this paper, we present a new networked system for efficient encrypted video delivery while preserving the benefits of in-network caching. As video chunks are encrypted before distribution, we first design a compact, efficient, yet encrypted video fingerprint index to empower the network with a fully controlled capability of locating the cached encrypted chunks for given encrypted requests. We then explain how to deploy the encrypted design in our proposed architecture and present a secure redundancy elimination protocol to enable fast video delivery via leveraging cached encrypted chunks. We further discuss the full support of cache management, adaptive video delivery, and video access control. Rigorous analysis and prototype evaluations demonstrate the security, efficiency, and effectiveness of the design.

As e-health technology continues to advance, health related multimedia data is being exponentially generated from healthcare monitoring devices and sensors. Coming with it are the challenges on how to efficiently acquire, index, and process such a huge amount of data for effective healthcare and related decision making, while respecting user's data privacy. In this paper, we propose a secure cloud-based framework for privacy-aware healthcare monitoring systems, which allows fast data acquisition and indexing with strong privacy assurance. For efficient data acquisition, we adopt compressive sensing for easy data sampling, compression, and recovery. We then focus on how to secure and fast index the resulting large amount of continuously generated compressed samples, with the goal to achieve secure selected retrieval over compressed storage. Among others, one particular challenge is the practical demand to cope with the incoming data samples in high acquisition rates. For that problem, we carefully exploit recent efforts on encrypted search, efficient content-based indexing techniques, and fine-grained locking algorithms, to design a novel encrypted index with high-performance customization. It achieves memory efficiency, provable security, as well as greatly improved building speed with nontrivial multithread support. Comprehensive evaluations on Amazon Cloud show that our encrypted design can securely index 1 billion compressed data samples within only 12 min, achieving a throughput of indexing almost 1.4 million encrypted samples per second. Accuracy and visual evaluation on a real healthcare dataset shows good quality of high-value retrieval and recovery over encrypted data samples.

As today's integrated circuit (IC) has easily involved millions and even billions of gates, known as very large-scale integration (VLSI), one natural trend is to move such prohibitive in-house design procedure to the low-cost public cloud. However, such a migration is also raising a challenging request on practical and privacy-preserving techniques to safeguard the sensitive IC design data, i.e., the Intellectual Property (IP). In this paper, we initiate the first study along the direction, and present a practical system for privacy-preserving IC timing analysis, which refers to an essential and expensive procedure via repeated evaluations of timing delays on a gate-level circuit. For privacy protection, our system leverages a key observation that many IP blocks are universally reused and shared across different IC designs, and thus only a small portion of critical IP blocks need to be protected. By carefully extracting such critical design data from the whole circuit, our system only outsources the non-critical data to the public cloud. As such "data splitting" does not readily facilitate correct timing analysis, we then develop specialized algorithms to enable the public cloud to take only the non-critical data and return intermediate results. Such results can later be integrated with critical design data by the local server for fast timing analysis. We also propose a heuristic algorithm to considerably reduce the bandwidth cost in the system. Through rigorous security analysis, we show our system is resilient to IC reverse engineering and protects both the critical IP gate-level design and functionality. We evaluate our system over large IC benchmarks with up to a million of gates to show the efficiency and effectiveness.

Multimedia contents, especially videos, are being exponentially generated today. Due to the limited local storage, people are willing to store the videos at the remote cloud media center for its low cost and scalable storage. However, videos may have to be encrypted before outsourcing for privacy concerns. For practical purposes, the cloud media center should also provide the deduplication functionality to eliminate the storage and bandwidth redundancy, and adaptively disseminate videos to heterogeneous networks and different devices to ensure the quality of service. In light of the observations, we present a secure architecture enabling the encrypted cloud media center. It builds on top of latest advancements on secure deduplication and video coding techniques, with fully functional system implementations on encrypted video deduplication and adaptive video dissemination services. Specifically, to support efficient adaptive dissemination, we utilize the scalable video coding (SVC) techniques and propose a tailored layer-level secure deduplication strategy to be compatible with the internal structure of SVC. Accordingly, we adopt a structure-compatible encryption mechanism and optimize the way how encrypted SVC videos are stored for fast retrieval and efficient dissemination. We thoroughly analyze the security strength of our system design with strong video protection. Furthermore, we give a prototype implementation with encrypted end-to-end deployment on Amazon cloud platform. Extensive experiments demonstrate the practicality of our system.

In storage outsourcing, highly correlated datasets can occur commonly, where the rich information buried in correlated data can be useful for many cloud data generation/dissemination services. In light of this, we propose to enable a secure and efficient cloud-assisted image sharing architecture for mobile devices, by leveraging outsourced encrypted image datasets with privacy assurance. Different from traditional image sharing, the proposed design aims to save the transmission cost from mobile clients, by directly utilizing outsourced correlated images to reproduce the image of interest inside the cloud for immediate dissemination. While the benefits are obvious, how to leverage the encrypted image datasets makes the problem particular challenging. To tackle the problem, we first propose a secure and efficient index design that allows the mobile client to securely find from the encrypted image datasets the candidate selection pertaining to the image of interest for sharing. We then design two specialized encryption mechanisms that support the secure image reproduction inside the cloud directly from the encrypted candidate selection. We formally analyze the security strength of the design. Our experiments show that up to 90% of the transmission cost at the mobile client can be saved, while achieving all service requirements and security guarantees.

Mobile advertising has become a key economic force with the growing mobile app markets. Targeted advertising has been attracting the attention of business and research communities because of its high return on investment. However, the fact that it requires continuous collection of user profiles raises severe privacy concerns. In this paper, we propose a secure and practical mobile advertising architecture, which aims to enable targeted advertising over encrypted mobile user profiles. The proposed idea is based upon a cryptographic primitive, i.e., private stream searching (PSS). We first propose a basic construction of secure targeted mobile advertising. Thorough analysis shows that our basic design achieves semantic security. After that, we introduce a number of techniques to make our design more efficient for mobile devices. Experimental results demonstrate the practicality of our proposed architecture.

The increasing popularity of images at social media sites is posing new opportunities for social discovery applications, i.e., suggesting new friends and discovering new social groups with similar interests via exploring images. To effectively handle the explosive growth of images involved in social discovery, one common trend for many emerging social media sites is to leverage the commercial public cloud as their robust backend data center. While extremely convenient, directly exposing content-rich images and the related social discovery results to the public cloud also raises new acute privacy concerns. In light of the observation, in this paper we propose a privacy-preserving social discovery service architecture based on encrypted images. As the core of such social discovery is to compare and quantify similar images, we first adopt the effective Bag-of-Words model to extract the "visual similarity content" of users' images into image profile vectors, and then model the problem as similarity retrieval of encrypted high-dimensional image profiles. To support fast and scalable similarity search over hundreds of thousands of encrypted images, we propose a secure and efficient indexing structure. The resulting design enables social media sites to obtain secure, practical, and accurate social discovery from the public cloud, without disclosing the encrypted image content. We formally prove the security and discuss further extensions on user image update and the compatibility with existing image sharing social functionalities. Extensive experiments on a large Flickr image dataset demonstrate the practical performance of the proposed design. Our qualitative social discovery results show consistency with human perception.

Level-sensitive transparent latches are widely used in high-performance sequential circuit designs. Under process variations, the timing of a transparently latched circuit will adapt random delays at runtime due to time borrowing. The central problem to determine the timing yield is to compute the probability of the presence of a positive cycle in the latest latch timing graph. Existing algorithms are either optimistic since cycles are omitted or require iterations that cannot be polynomially bounded. In this paper, we present the first algorithm to compute such probability based on block-based statistical timing analysis that, first, covers all cycles through a structural graph traversal, and second, terminates within a polynomial number of statistical ¿sum¿ and ¿max¿ operations. Experimental results confirm that the proposed approach is effective and efficient.