Ensuring web accessibility is crucial for advancing social welfare, justice, and equality in digital spaces, yet the vast majority of website user interfaces remain non-compliant, due in part to the resource-intensive and unscalable nature of current auditing practices. While WCAG-EM offers a structured methodology for site-wise conformance evaluation, it involves great human efforts and lacks practical support for execution at scale. In this work, we present an auditing framework, AAA, which operationalizes WCAG-EM through a human-AI partnership model. AAA is anchored by two key innovations: GRASP, a graph-based multimodal sampling method that ensures representative page coverage via learned embeddings of visual, textual, and relational cues; and MaC, a multimodal large language model-based copilot strategy that supports auditors through cross-modal reasoning and intelligent assistance in high-effort tasks. Together, these components enable scalable, end-to-end web accessibility auditing, empowering human auditors with AI-enhanced assistance for real-world impact. We further contribute four novel datasets designed for benchmarking core stages of the audit pipeline. Extensive experiments demonstrate the effectiveness of our methods, providing insights that small-scale language models can serve as capable experts when fine-tuned.

Realistic background traffic is critical to the simulation platforms for autonomous driving (AD) testing. Given that most vehicles in reality are driven by human beings, introducing human driving (HD) vehicles to the background traffic is necessary to be able to discover more problems of the tested AD vehicle in the simulation stage. However, existing methods rely on ad-hoc rules or data-driven training to mimic partial human driver behaviors, which are not comprehensive and lack transparency. In this work, we design a smart human driving vehicle simulator HDSim which is empowered by cognitively inspired modeling and AI models. HDSim enables diverse, realistic, and scalable HD traffic simulation on AD testing platforms like CARLA in a non-intrusive manner. There are two novel components in HDSim. First, we introduce a driver model to guide the generation of diverse human driving styles by using different combinations of latent cognitive factors in a hierarchy. Second, we design a Perception-Mediated Behavior Influence (PMBI) mechanism to use LLM-assisted perceptual transformations to indirectly fuse driving actions with driving styles. Experiments show that HDSim traffic can help simulation platforms like CARLA to reveal 68% more failures of tested AD vehicles, and the explainability of reported accidents is also improved.

Large Language Model (LLM) inference has emerged as a fundamental paradigm, however, variations in output length cause severe workload imbalance in the decode phase, particularly for long-output reasoning tasks. Existing systems, such as PD disaggregation architectures, rely on static prefill-to-decode scheduling, which often results in SLO violations and OOM failures under evolving decode workloads. In this paper, we propose STAR, a decode rescheduling system powered by length prediction to anticipate future workloads. Our core contributions include: (1) A lightweight and continuous LLM-native prediction method that leverages LLM hidden state to model remaining generation length with high precision (reducing MAE by 49.42%) and low overhead (cutting predictor parameters by 93.28%); (2) A rescheduling solution in decode phase with a dynamic balancing mechanism that integrates current and predicted workloads, reducing P99 TPOT by 75.1% and achieving 2.63 × higher goodput.

The widespread adoption of online Large Language Models (LLMs) raises considerable privacy concerns, as prompts may inadvertently contain sensitive information exposed to LLM service providers. Limited by high computational costs, reduced response utility, and excessive system modifications, previous works based on local deployment, embedding perturbation, and homomorphic encryption are not feasible for online prompt-based LLM services. To address these issues, we introduce ProSan (Prompt Privacy Sanitizer), an end-to-end method for prompt privacy protection that generates prompts with task-irrelevant privacy removed, while preserving both utility and readability. It can also be seamlessly integrated into the online LLM service pipeline. To achieve high utility and contextual privacy, ProSan flexibly adjusts its protection targets and strength based on the importance of the words and the privacy leakage risk of the prompts. Additionally, ProSan is capable of adapting to diverse computational resource conditions, ensuring privacy protection for low-resource users. Our experiments demonstrate that ProSan effectively removes sensitive information across various tasks, including question answering, text summarization, and code generation, with minimal reduction in task performance.

Mobile video calls play an essential role in our daily lives. However, in bandwidth-limited scenarios (e.g., inadequate cellular coverage, congested satellite links, and metered connections), users often experience poor quality of experience (QoE) during video calls. While recent advances in deep learning have demonstrated significant improvements in video compression over traditional methods, existing approaches are ill-suited for bidirectional video streaming on smartphones. The primary challenge lies in simultaneously achieving high video quality, computational and bandwidth efficiency, and practical usability on constrained mobile devices. In this work, we present SlimFit-Gens, the first practical video calling system for smartphones capable of delivering real-time 480p video at as low as 30 kbps. SlimFit-Gens addresses the challenge with joint algorithm and system-level optimizations. The core technique is a fine-grained model personalization design tailored for mobile video calling, enabling high-fidelity video generation at low model complexity. SlimFit-Gens achieves effective personalized adaptation through a novel two-stage personalization mechanism working upon an optimized model architecture. It also incorporates a privacy preserving, resource-efficient system design, featuring TEE-based (e.g., Confidential VM/NVIDIA Confidential Computing) fine-tuning on the server side and heterogeneity-aware inference on the device side. We implement SlimFit-Gens on four commercial off-the-shelf (COTS) smartphones with different system-on-chip (SoC) configurations and conduct extensive evaluations. Compared to prior work, SlimFit-Gens simultaneously improves generation quality with a 0.09-0.12 reduction in LPIPS and system efficiency through a 1.6-1.8× increase in video frame rate.

Yu Lin, Ruining Yang, Yunlong Mao, Qizhi Zhang, Jue Hong, Quanwei Cai, Ye Wu, Huiqi Liu, Zhiyu Chen, Bing Duan, Sheng Zhong. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Electromagnetic (EM) covert channels pose significant threats to computer and communications security in air-gapped networks. Previous works exploit EM radiation from various components (e.g., video cables, memory buses, CPUs) to secretly send sensitive information. These approaches typically require the attacker to deploy highly specialized receivers near the victim, which limits their real-world impact. This paper reports a new EM covert channel, TEMPEST-LoRa, that builds on Cross-Technology Covert Communication (CTCC), which could allow attackers to covertly transmit EM-modulated secret data from air-gapped networks to widely deployed operational LoRa receivers from afar. We reveal the potential risk and demonstrate the feasibility of CTCC by tackling practical challenges involved in manipulating video cables to precisely generate the EM leakage that could readily be received by third-party commercial LoRa nodes/gateways. Experiment results show that attackers can reliably decode secret data modulated by the EM leakage from a video cable at a maximum distance of 87.5m or a rate of 21.6 kbps. We note that the secret data transmission can be performed with monitors turned off (therefore covertly).

Estimating the difficulty of input questions as perceived by large language models (LLMs) is essential for accurate performance evaluation and adaptive inference.Existing methods typically rely on repeated response sampling, auxiliary models, or fine-tuning the target model itself, which may incur substantial computational costs or compromise generality.In this paper, we propose a novel approach for difficulty estimation that leverages only the hidden representations produced by the target LLM.We model the token-level generation process as a Markov chain and define a value function to estimate the expected output quality given any hidden state.This allows for efficient and accurate difficulty estimation based solely on the initial hidden state, without generating any output tokens.Extensive experiments across both textual and multimodal tasks demonstrate that our method consistently outperforms existing baselines in difficulty estimation.Moreover, we apply our difficulty estimates to guide adaptive reasoning strategies, including Self-Consistency, Best-of-N, and Self-Refine, achieving higher inference efficiency with fewer generated tokens.

Collaborative driving has emerged as a promising approach to improve vehicle positioning accuracy. However, existing collaborative driving systems often face substantial privacy leakage issues. In this paper, we propose EPPCoDrive, a novel, efficient, and privacy-preserving collaborative driving protocol tailored for real-world vehicular networks. Specifically, EPPCoDrive leverages lightweight cryptographic mechanisms to protect vehicle identifiers and locations against potential adversaries. Experimental results show that our system preserves a positioning accuracy comparable to traditional, non-privacy solutions, while introducing minimal computational overhead.

Traditional image steganography focuses on transmitting the secret message over lossless channels. However, those schemes are vulnerable to image scaling attacks, especially image downsampling operations. Nowadays, substantial achievements have been made on robust steganography against JPEG compression and general scaling attacks. Existing methods can transmit the secret message through compressions with different scale factors, but they rely highly on pre-detecting the lossy channel’s scale factor and modifying their strategy accordingly. Subsequently, they are unable to resist alterable compressions or survive variable channels. This paper proposes a robust steganography system that operates without pre-detection, effectively solving the above problem when the lossy channel adopts the Nearest-Neighbor Interpolation to scale images. The system demonstrates resistance against multi-scale scaling attacks, enabling flawless data extraction while maintaining a considerably high embedding rate.

Triangle counting is essential for analyzing network structures and optimizing recommendation systems, yet it can lead to privacy breaches if individual data is not adequately protected during the analysis. Differential privacy has become a widely adopted standard to safeguard personal privacy. Current research mainly focuses on central and local models, which differ in applicability and performance. The central model cannot be applied when graph data is distributed among multiple parties without a trusted central server, while the local model provides unsatisfactory performance. In this paper, we explore a two-party scenario where each party holds private information about a group of users and is not allowed to disclose this information to the other party. We have proposed a scheme called HTTC-DPk, which ensures both differential privacy and k-anonymity, and is better suited to the constraints of the two-party setting compared to both central and local models. Our method integrates the noisy maximum degree computation for both intra-party and inter-party edges, along with the differentially private inter-party triangle counting based on two-party interactions. Additionally, we introduce an enhanced scheme HTTC-DPk* that strikes a balance between accuracy and communication costs, particularly suitable for large-scale graphs. We have provided comprehensive theoretical proof of our scheme's privacy and demonstrated through extensive experiments that our approach performs well under various cases.

Pre-trained Language Models (PLM) have enabled a cost-effective approach to handling various downstream applications via Parameter-Efficient-Fine-Tuning (PEFT) techniques. In this context, service providers have introduced a popular fine-tuning-based product service known as Model-as-a-Service (MaaS). This service offers users access to extensive PLMs and training resources. With MaaS, users can fine-tune, deploy, and utilize their customized models seamlessly, leveraging a one-stop platform that allows them to work with their private datasets efficiently. However, this service paradigm has recently been exposed to the possibility of leaking user private data. To this end, we identify the data privacy leakage risks in MaaS-based PEFT and propose a Split-and-Privatize (SAP) framework, mitigating the privacy leakage by integrating split learning and differential privacy into MaaS PEFT. Furthermore, we propose Contributing-Token-Identification (CTI), a novel method to balance model utility degradation and privacy leakage. As a result, the proposed framework is comprehensively evaluated, demonstrating a 65% improvement in empirical privacy with only a 1% degradation in model performance on the Stanford Sentiment Treebank dataset, outperforming existing state-of-the-art baselines.

The frequent impacts and serious damages of cyberattacks raise an urgent need for robust network intrusion detection amidst escalating cyber threats. Traditional signature-based methods often fail to detect novel attacks, necessitating more adaptive anomaly-based approaches. To address the challenge of effective network traffic modeling, we propose a novel method named E-BiGraphSAGE, which captures intricate relationships between source and destination addresses of network traffic activities in a bipartite graph. The proposed method comprises three main components: flow-centric bipartite graph construction, backtracking bipartite graph learning, and vertex splitting strategy. Firstly, a flow-centric bipartite graph is constructed to better represent the flow-based dataset, effectively capturing the crucial information of network flows. Secondly, a backtracking bipartite graph learning model is introduced to learn the representations of heterogeneous vertices within this flow-centric bipartite graph. Furthermore, a vertex splitting strategy is implemented to reduce the computational complexity of the method. Empirical evaluations on four benchmark NIDS datasets demonstrate that E-BiGraphSAGE outperforms state-of-the-art methods in terms of detection accuracy and robustness, offering a more effective solution for network intrusion detection.

In recent years, there has been significant progress in extending convolutional neural networks to graph structures, leading to the development of various graph convolution operators for performing convolution in the graph domain. Pooling, another essential operation, plays a crucial role in progressively reducing the spatial dimensions of representations, facilitating the creation of hierarchical representations and minimizing the number of parameters. However, research on pooling operators specifically designed for graphs remains limited. Some existing methods focus on selecting a subset of nodes based on scores derived from trainable projection vectors or graph convolution layers. Unfortunately, these approaches often overlook the locality of nodes, which can result in substantial information loss from the original graphs. To address this challenge, we introduce a learnable graph local pooling method, termed LocalPool. This method first employs graph neural network operations to learn the importance scores for each node. To preserve the structural information of the graph, we implement an efficient vertex-weighted graph cut to cluster adjacent nodes into supernodes within coarsened graphs. The features of these supernodes are then generated by weighting the node features in their respective clusters according to the learned importance scores. Additionally, the adjacency matrix of the coarsened graphs is derived from the clustering information of the original graphs. We evaluate our method against existing graph pooling techniques on graph classification benchmarks, maintaining the same network architecture. The experimental results demonstrate that LocalPool consistently outperforms existing methods.

Text-to-image synthesis has become highly popular for generating realistic and stylized images, often requiring fine-tuning generative models with domain-specific datasets for specialized tasks. However, these valuable datasets face risks of unauthorized usage and unapproved sharing, compromising the rights of the owners. We address the issue of dataset abuse during the fine-tuning of Stable Diffusion (SD) models for text-to-image (T2I) synthesis. We present a dataset watermarking framework designed to detect unauthorized usage and trace data leaks. Experiments demonstrate the framework's effectiveness, minimal impact on the dataset (only 2% of the data required to be modified for high detection accuracy), and ability to trace data leaks. Our results also highlight the transferability and robustness of the framework, proving its practical applicability in detecting dataset abuse. The code is available: https://github.com/inmbzpmdqv/treidzxyq.git

Large Language Models (LLMs) have become integral to automated code analysis, enabling tasks such as vulnerability detection and code comprehension. However, their integration introduces novel attack surfaces. In this paper, we identify and investigate a new class of prompt-based attacks, termed Copy-Guided Attacks (CGA), which exploit the inherent copying tendencies of reasoning-capable LLMs. By injecting carefully crafted triggers into external code snippets, adversaries can induce the model to replicate malicious content during inference. This behavior enables two classes of vulnerabilities: inference length manipulation, where the model generates abnormally short or excessively long reasoning traces; and inference result manipulation, where the model produces misleading or incorrect conclusions. We formalize CGA as an optimization problem and propose a gradient-based approach to synthesize effective triggers. Empirical evaluation on state-of-the-art reasoning LLMs shows that CGA reliably induces infinite loops, premature termination, false refusals, and semantic distortions in code analysis tasks. While highly effective in targeted settings, we observe challenges in generalizing CGA across diverse prompts due to computational constraints, posing an open question for future research. Our findings expose a critical yet underexplored vulnerability in LLM-powered development pipelines and call for urgent advances in prompt-level defense mechanisms.

Unmanned sanitation vehicles (USVs) promise cleaner cities, yet efficiently coordinating multiple USVs in large urban areas remains challenging due to constraints such as limited waste capacity and battery life. In this demo, we present MRTC, a multi-robot task cooperation system. First, Dynamic Task Assignment employs an Actor-Critic policy within a Markov decision framework to allocate cleaning tasks and decide the required number of USVs. Second, Single-USV Path Planning refines each route via a fast two-layer iterative search. Over an eight-month real-world deployment in three urban testbeds, our MRTC system markedly improved cleaning efficiency while lowering operating costs. Operating over a combined 10,775 km of routes per month, the system achieved average monthly savings of 20,575 kWh of energy and 2,744 labour hours. A demonstration video is available at https://llq978.github.io/Demo/.

The Louvain algorithm is one of the most popular algorithms for community detection. Observing that existing implementations suffer from inaccurate pruning and inefficient intermediate state management, we introduce GALA, GPU-Accelerated Louvain Algorithm, which incorporates two key innovations. The first innovation is a novel modularity gain-based pruning strategy, supported by rigorous theoretical guarantees of optimality and able to reduce up to 76% of vertices as well as their corresponding computations. To take advantage of the memory hierarchy and parallelism of GPUs, the second innovation is workload-aware kernels, featuring a shuffle-based kernel founded on the warp-level primitives for exchange states and a hash-based kernel that prioritizes shared memory in hashtable design. GALA further scales to multiple GPUs by minimizing the synchronization overhead between GPUs through a dense-sparse synchronization strategy. We evaluate the performance of GALA through theoretical analysis and practical experiments on various real-world graphs. The experimental results confirm that GALA significantly improves the performance of the parallel Louvain algorithm on GPUs, surpassing state-of-the-art solutions by 6× on average.

LLM-based vulnerability auditors (e.g., GitHub Copilot) represent a significant advancement in automated code analysis, offering precise detection of security vulnerabilities. This paper explores the potential to circumvent LLM-based vulnerability auditors by diverting their focus, decided by the LLM attention mechanism, away from real vulnerable code segments. In these LLM-based vulnerability auditors, the attention mechanism is supposed to focus on potentially vulnerable code sections to identify security issues. Our approach introduces high-attention code snippets (code fragments designed to draw focus) into the codebase under review. By strategically diverting the model's focus away from actual vulnerabilities, this technique effectively “blinds” the LLM, resulting in missed detections. To scale this approach, we present Crazy-Ivan11Source code, dataset and attack results are available at https://github.com/oxygen-hunter/Flashboom., an automated system that identifies and seamlessly integrates high-attention code snippets, shifting focus away from genuine vulnerabilities to decoy functions. Through systematic function-level prioritization and refinement, Crazy-Ivan optimizes the blinding effect, producing the Flashboom that can reduce the model's capacity to detect true security risks. Our evaluation underscores the effectiveness of Flashboom, achieving blinding success rates of up to 96.3% on CodeLlama and 83.05% on Gemma, with notable cross-model transferability and applicability across multiple programming languages. In a case study with GitHub Copilot, Flashboom led the tool to overlook a critical blockchain vulnerability, underscoring the security implications of such attention-diverting attacks and the risks inherent in relying solely on LLM-based automated auditing systems. We have reported our findings to the respective LLM-based code auditor vendors, who have acknowledged the issues and are currently working on fixes.

Payment channel networks (PCNs) offer a crucial solution to the scalability challenges of blockchain-based transaction systems. However, most existing PCN routing protocols employ a “guess-and-check” approach, which undermines their transaction success rate and efficiency. In this paper, we propose a routing protocol named RCS, based on a novel “Refined Confirm-and-Send” approach. Utilizing PCN topology statistics, RCS performs a refined probing of possible transaction paths and verifies whether a path has sufficient available balance before executing the transaction through it. This method effectively improves the transaction success rate while maintaining restrained overhead. Additionally, to address users' privacy concerns, we design a privacy-preserving version of RCS, named RCS+. RCS+ uses secure comparisons to identify paths with sufficient funds without disclosing channel balances or transaction amounts. Extensive simulations with real-world and synthetic datasets demonstrate that RCS and RCS+ outperform existing state-of-the-art protocols. RCS and RCS+ achieve a$\mathbf{1 0 \%}$higher transaction success rate compared to the Shortest Path approach, which serves as the core of Lightning Network's current routing mechanism. In terms of overhead, RCS maintains the lowest cost among all tested protocols, e.g., only 20 % of the Flash protocol. While RCS+ incurs marginally higher overhead due to its enhanced privacy guarantees, its cost remains just 30 % of Flash's overhead. Furthermore, RCS/RCS+ exhibits robust adaptability to dynamic changes in PCN topologies, ensuring scalability as the network evolves.

Graph data analysis, particularly local triangle counting, plays a pivotal role in deciphering complex relationships within graph data. This method is invaluable across diverse fields such as social networks, transportation, and cybersecurity. However, this process often involves handling sensitive information, necessitating that the relationship between any two nodes is considered private. Differential privacy (DP) is a formal model to address privacy concerns and can be categorized into two types: the central DP (CDP) model, which achieves better result accuracy, and the local DP (LDP) model, which does not assume a trusted server. To bridge the gap between the two models, we propose Sectric, a server-aided crypto-assisted local triangle counting protocol, in this paper. It can achieve the same result accuracy with the same privacy budget as the CDP model without assuming a trusted server. Sectric also explores a new approach in crypto-assisted graph data analysis algorithms that represents a node's neighbors using a set instead of an adjacency vector, and successfully achieves higher efficiency compared to other crypto-assisted solutions. We also conduct theoretical and empirical evaluations to demonstrate that Sectric achieves the design principles.

With the development of cryptographic tools such as Fully Homomorphic Encryption (FHE) and secure Multiparty Computation (MPC), privacy-preserving Machine Learning as a Service (MLaaS) has gained attractiveness for its security when it comes to utilizing cross-domain data. However, cryptographic tools are characterized by huge overhead, which results in the MLaaS quality being unbearably degraded, especially for latency-sensitive MLaaS applications. In this paper, we focus on the problem of low-latency inference associated with MLaaS and propose CrossNet, a Privacy-preserving Neural Network Inference (PPNI) framework based on FHE, for applications with limited client-side computational and communication resources. CrossNet performs model transformations on neural networks so that they can be evaluated in an FHE-friendly manner. Model transformation introduces limited interactions between client and server, thus restricting inference latency. In addition, CrossNet includes a series of layer constructions where elaborate encoding forms and computational orders are designed to further reduce the overhead of transformed layers. CrossNet outperforms the existing FHE-based frameworks by 4x efficiency and reduces nearly 30% inference latency on ResNet-50 in a resource-limited setting.

The Payment Channel Network (PCN) stands out as one of the most promising technologies for scaling blockchain-based cryptocurrencies. However, a noteworthy challenge arises during the utilization of PCNs, where a substantial portion of payment channels gradually becomes exhausted, leading to a reduction in the overall availability of PCNs. This issue is crucial in the context of blockchain off-chain PCNs and warrants a comprehensive investigation. In this paper, we introduce the problem of optimal recover and propose OPtimal REcovering protocols, denoted asOPREandOPRE+, to address this challenge. The protocols target at recovering the optimal number of nearly exhausted channels in the PCN. OPRE provides a basic solution, and OPRE+ is an augmentation which provides a more efficient and effective solution. Furthermore, to address users’ privacy concerns, we propose privacy-preserving versions of the protocols, ensuring that users’ balance on payment channels remains undisclosed during the execution of the protocols. Beyond the theoretical design and analysis, we implement these protocols and conduct experimental evaluations to assess their performance. The results affirm that our protocols exhibit efficiency and effectiveness in significantly improving the availability of PCNs.

Recently, watchtowers emerge as a critical service within payment channel networks (PCNs). Existing payment channels necessitate that channel owners periodically monitor the blockchain to ensure their fund security, or enlist the watchtower services for this task. Presently, watchtower proposals mandate the implementer to provide collateral as a safeguard against its collusion with potential adversaries. However, this collateral substantially inflates the implementation costs, consequently leading to higher service fees for users. Furthermore, most watchtowers are typically operated by a third-party entity, creating a single point of failure. To ameliorate the status quo, we propose a novel approach where PCN nodes collaboratively implement a watchtower system named “SilenTower.” Our proposal is rooted in the fundamental principles of blockchain systems, emphasizing maintenance by a community with an honest majority. SilenTower’s security no longer relies on collateral but rather on the inherent difficulty of a large proportion of collusion. SilenTower also allows inaccessible participants and thus obtains robustness. Through a rigorous theoretical analysis, we demonstrate that participants’ optimal strategy is to remain accessible and faithfully adhere to the SilenTower protocol. Furthermore, we assess the practical performance of SilenTower through comprehensive benchmarking, and the results reveal that it introduces lightweight overheads.

Designing an efficient and secure collaborative SQL analysis system that supports large-scale dataset inputs is a very challenging task. In this paper, we present FedQuery, an MPC-based solution for efficient and secure collaborative analysis that is able to handle billion-scale dataset inputs. FedQuery introduces novel designs in its system architecture, the underlying MPC primitives, and oblivious SQL operators as well as their combinations, significantly reducing communication and computation overhead. Comprehensive experiments on real-world datasets show that FedQuery achieves large performance improvements over state-of-the-art baselines at both the operator and query levels. Additionally, it can handle complex SQL queries on datasets up to ten billion entries in less than 14 hours.

Dynamic Class Loading (DCL) is a legitimate technique extensively used by Android developers to incorporate additional functionalities into applications at runtime. However, adversaries can exploit DCL as a stealthy obfuscation technique to dynamically load malicious code and evade detection. While prior studies have analyzed typical DCL-based obfuscation and the attacks it enables—such as identifying payloads on storage, inspecting DCL-related APIs, or profiling dynamic behaviors—existing solutions remain insufficient against increasingly evasive DCL threats. In this paper, we propose UTRDCL, a novel stealthy obfuscation technique that leverages system APIs instead of conventional DCL-related APIs, and employs an automated footprint cleanup strategy to minimize runtime traces. Based on UTRDCL, we construct three real-world attack instances by embedding it into existing malware and benign applications, demonstrating how it can be used to evade detection. To counter such threats, we design and implement a lightweight defense mechanism by patching a previously overlooked vulnerability in the Android system that UTRDCL exploits in this specific context. This system-level mitigation closes the attack surface leveraged by UTRDCL, offering a more fundamental defense than behavioral detection. Extensive experiments show that attacks leveraging UTRDCL can evade 11 state-of-the-art malware detectors from open-source, academic, and commercial sources. We further validate our defense mechanism on real devices, demonstrating its effectiveness in preventing UTRDCL-based attacks without introducing noticeable overhead. Our proof-of-concept of UTRDCL and its defense is publicly available1.

The emerging virtual machine-based Android packers render existing unpacking techniques ineffective. The state-of-the-art unpacker falls short because it relies on unreliable heuristics and manually crafted semantic models. Hence, it cannot precisely recover app semantics necessary for malware detection. In this paper, we proposeDeepVMUnProtect, a deep learning-based approach to automatically and accurately capture the semantics of VM-packed code, so as to facilitate semantic-based Android malware classification. Experiments have shown thatDeepVMUnProtectoutperforms the state-of-the-art tool on recovering opcode semantics in Qihoo(58.3%), Baidu(47.5%) and NMMP (58.8%) respectively, and can enable semantics-aware malware detection which prior work fails to do.

Frequent itemset mining is an essential task in data analysis. Therefore, it is crucial to design practical methods for privacy-preserving frequent itemset mining, enabling private data analysis. For two-party data analysis tasks, each party possesses its portion of the data and is reluctant to share the data with the other. While secure computation can enable two-party frequent itemset mining, the output of exact top-$k$itemsets may still leave the adversary a chance to infer the sensitive information. Differential privacy has been utilized in various data analysis tasks to safeguard participating individuals. However, addressing how to ensure differential privacy for two-party frequent itemset mining remains unexplored. To prevent each party’s data from being leaked to the other while achieving differential privacy for releasing the output, this paper investigates the problem of differentially private two-party frequent itemset mining. We first propose a practical method that can efficiently select the frequent items of the union of two confidential databases in a differentially private way without the need to combine all elements. Then we extend this technique for general frequent itemset mining. Extensive experiments were conducted on real-world datasets, and the results show that the proposed method can achieve satisfactory utility with affordable overheads.

Local differential privacy (LDP) provides a way for an untrusted data collector to aggregate users' data without violating their privacy. Various privacy-preserving data analysis tasks have been studied under the protection of LDP, such as frequency estimation, frequent itemset mining, and machine learning. Despite its privacy-preserving properties, recent research has demonstrated the vulnerability of certain LDP protocols to data poisoning attacks. However, existing data poisoning attacks are focused on basic statistics under LDP, such as frequency estimation and mean/variance estimation. As an important data analysis task, the security of LDP frequent itemset mining has yet to be thoroughly examined. In this paper, we aim to address this issue by presenting novel and practical data poisoning attacks against LDP frequent itemset mining protocols. By introducing a unified attack framework with composable attack operations, our data poisoning attack can successfully manipulate the state-of-the-art LDP frequent itemset mining protocols and has the potential to be adapted to other protocols with similar structures. We conduct extensive experiments on three datasets to compare the proposed attack with four baseline attacks. The results demonstrate the severity of the threat and the effectiveness of the proposed attack.

In this paper, we address the critical concerns related to dataset privacy in the context of k-means clustering publishing within a distributed dimension setting. By leveraging differential privacy mechanisms, we propose a novel framework that integrates a differentially private classifier, constructed through voting based on raw clustering results, and an enhanced generative adversarial network (GAN) simulating the classifier’s behavior in inferring class labels for a public dataset. Our approach generates synthetic clustering results that mimic real outcomes in classification tasks, ensuring differential privacy and minimizing noise. Our contributions include a comprehensive exploration of privacy issues, the introduction of a novel privacy-preserving k-means clustering framework, and theoretical analyses demonstrating sensitivity and differential privacy guarantees. Evaluation on the MNIST dataset demonstrates the effectiveness of the framework, achieving 82.22% accuracy with a (10.48, 10−9)-differential-privacy guarantee, compared to 83.45% accuracy without privacy-preserving.

Three-dimensional (3D) scene reconstruction, particularly from monocular videos, is a significant challenge in large-scale scenarios due to difficulty handling varying object sizes and high computational resource needs. This paper introduces MuSR, a novel multi-scale reconstruction method addressing these issues. MuSR features a dynamic multi-resolution spatial structure that adaptively adjusts voxel resolution for objects of different sizes to improve reconstruction quality. MuSR also employs a block-based sparse 3D data structure and hardware resource management strategy to reduce GPU memory usage while maintaining efficient reconstruction. Evaluated on ScanNet and 7-Scenes datasets, as well as real-world scenes, MuSR outperforms state-of-the-art methods in terms of efficiency, completeness, and geometric shape reconstruction, proving its applicability in practical multi-scale 3D reconstructions.

With growing network service demands, rational and efficient multi-domain resource allocation is paramount. Research aims to develop intelligent Traffic Engineering (TE) algorithms that can dynamically allocate resources, adapt to changing conditions, and meet user needs. TE algorithms based on Software-Defined Networking (SDN) have proven effective for this goal by leveraging the centralized control plane and programmable data plane of SDN. This enables flexible and dynamic optimization of routing and resource allocation across multiple domains to meet traffic demands. However, domains operated by different service providers may exhibit non-cooperative behavior due to conflicts of interest and competition. Some domains may act selfishly by hiding bandwidth or exaggerating inter-domain requests to reserve more resources for themselves. This complicates TE design as algorithms can no longer assume universal cooperation in multi-domain networks. Game theory provides a framework to model competition between domains through behaviors like request forwarding, dropping and study cooperation strategies. This paper presents GameTE, a game-theoretic distributed TE algorithm for multi-domain SDN environments without trusted relationships between domains. By incorporating incentives and punishments, our algorithm suppresses selfish behaviors and promotes efficient resource utilization. Evaluation results demonstrate that GameTE is effective in curbing deception, enhancing resource sharing between domains, and improving overall network performance compared to baseline schemes.

3D reconstruction based on monocular videos has attracted wide attention, and existing reconstruction methods usually work in a reconstruction-after-scanning manner. However, these methods suffer from insufficient data collection problems due to the lack of effective guidance for users during the scanning process, which affects reconstruction quality. We propose VIDAR, which visually guides users with the streaming incremental reconstructed mesh in data collection for monocular 3D reconstruction. We propose an incremental mesh extraction algorithm to achieve lossless fusion of streaming incremental mesh data via slice-style management for guidance quality. We also design an incremental mesh rendering algorithm to achieve precise memory reallocation by updating the buffer in a fill-in-the-blank pattern for guidance efficiency. Besides, we introduce several optimizations on data transmission and human-computer interaction to improve the overall system performance. The experiment results on real-world scenes show that VIDAR efficiently delivers high-quality visual guidance and outperforms the non-interactive data collection methods for scene reconstruction.

In the federated learning (FL) process, since the data held by each participant is different, it is necessary to figure out which participant has a higher contribution to the model performance. Effective contribution assessment can help motivate data owners to participate in the FL training. Research works in this field can be divided into two directions based on whether a validation dataset is required. Validation-based methods need to use representative validation data to measure the model accuracy, which is difficult to obtain in practical FL scenarios. Existing validation-free methods assess the contribution based on the parameters and gradients of local models and the global model in a single training round, which is easily compromised by the stochasticity of model training. In this work, we propose CoAst, a practical method to assess the FL participants' contribution without access to any validation data. The core idea of CoAst involves two aspects: one is to only count the most important part of model parameters through a weights quantization, and the other is a cross-round valuation based on the similarity between the current local parameters and the global parameter updates in several subsequent communication rounds. Extensive experiments show that CoAst has comparable assessment reliability to existing validation-based methods and outperforms existing validation-free methods.

Federated learning (FL), as a distributed training paradigm, has drawn great attention from both academia and industry. Recently, privacy and security concerns have been raised for FL. Despite many efforts to protect privacy and security, an FL framework that can systematically provide privacy and security guarantees is lacking. In this work, we present TTFL, a trustworthy FL framework in practice to defend the security and privacy issues based on Arm Confidential Compute Architecture (CCA). TTFL has two core designs. (1) It achieves a high-availability privacy protection based on flexible Trusted Execution Environments (TEEs). It leverages the resource-rich and conveniently accessed features of the latest TEE on Arm CCA, combined with our TEE secure interconnection design, to enable the whole FL process performed in distributed TEEs, which efficiently protects parameter confidentiality and protocol integrity. (2) It achieves effective security protection by proposing an effective poisoning-resisted secure aggregation scheme and protecting it within TEE. The new proposed secure aggregation combines the advantages of existing defenses and is placed in the flexible TEE to ensure a secure, effective, and non-bypassable aggregation procedure. We implement a prototype of TTFL and evaluate it regarding security, privacy, and system performance. Evaluation results show that TTFL can comprehensively and efficiently address the main privacy and security threats in FL. For instance, compared with previous work, it improves the model accuracy by 1.9% and reduces the attack success rate by 79.7% on the CIFAR-10 dataset with only about 19.8% training time overhead.

Exploiting participants’ data without knowing them is the center topic of secure mobile sensing data aggregation. In this article, we study how to improve existing protocols for computing the minimum or$k$-th minimum of all participants’ data in a privacy-preserving manner. Existing protocols for these two computations require relatively high communication cost or frequent interactions, which leads to unbearable time consumption when the network delay is high. We improve the min computation protocol proposed by Zhang et al. 2017, cutting down on its need for interactions and thus making it perform better in terms of efficiency. We also propose a new protocol as a secure substitute of the Bit-choosing Algorithm designed by Yu et al. 2018. It helps participants generate a secret permutation, which will be further used in the$k$-th min computation protocol to achieve higher accuracy and efficiency. Theoretical analyses are done to help predict and understand our new protocols’ performances, and later evaluations show that both these new protocols perform notably better in comparison to the existing ones.

Conventional crowdsourcing platforms primarily rely on a central server as the broker for information exchange. Although many efforts have been made, centralized platforms are still vulnerable to underlying security issues, such as an untrusted central server and single-point failure. Fortunately, blockchain has emerged as an alternative infrastructure for building crowdsourcing platforms. Many excellent designs of blockchain-based decentralized crowdsourcing (BDCS) solutions have been proposed. Benefiting from blockchain, BDCS can provide fascinating features, like tampering resistance and anonymity. However, a new attack surface appears in BDCS. Recently, a new attack against BDCS named solution probing attack has been identified. The solution-probing adversary can take advantage of the anonymity of BDCS to probe valid solutions using a generative model. Due to the transparency of blockchain transactions, the probing attack is effective even if solutions are encrypted. Nevertheless, we find transaction-mixing techniques effective in defending against probing attacks. In this paper, we introduce the solution probing attack and an improved variant, which can attack coin mixing-based BDCS. We evaluate probing attacks on large-scale crowdsourcing tasks. Experimental results show that the adversary is capable of deceiving BDCS with a limited number of probing, even if the BDCS is protected by solution encryption and coin mixing techniques.

Model publishing and deployment are essential for artificial intelligence applications. A major challenge in model publishing is efficiently distributing the models in a scalable way without violating the privacy of sensitive data. With the wide adoption of machine learning techniques, the privacy concern has also drawn much attraction. Differential privacy has become an important notion for privacy protection and is popular in private learning. However, it may bring much accuracy loss to fulfill data privacy. In addition, the private models are also hard to train in terms of convergence, which makes the existing approaches not scalable for private model publishing. This paper proposes a model publishing framework that provides a novel way to train privacy-preserving machine learning models with fast convergence and a lower privacy budget. By incorporating the concept of iterative machine teaching and the techniques in differential privacy, we have explored a way to privately select more suitable examples in the training process for achieving good accuracy with fewer iterations. Our analysis shows the privacy and convergence performance of the proposed method, and extensive experiments have been performed on real-world datasets to demonstrate its effectiveness.

Vertical federated learning (VFL) is being used more and more widely in industry. One of its most common application scenarios is a two-party setting: a participant (i.e., the host), who exclusively owns the labels but possesses insufficient number of features, wants to improve its model performance by combining features from another participant (i.e., the client) of a different business group. The best deep ML architecture suits for this scenario is considered to be Split Neural Network (SplitNN), in which each participant runs a self-defined bottom model to learn the hidden representations (i.e., the local embeddings) of its local data and then forwards them to the host, who runs a top model to aggregate both the local embeddings to produce the final predicts. In this paper, we assume the client is malicious and demonstrate that she/he could inject a stealthy backdoor into the top model during the training to misclassify any sample to a pre-selected target class with a high probability by just replacing its local embedding with a special trigger vector regardless of the host-side embedding. This task is non-trivial because existing data poison attacks for backdoor injection in traditional models usually require to modify the labels of a set of trigger-tagged samples of non-target classes, which is impossible here as the client has no rights to access or modify the labels exclusively owned by the host. Targeting this challenge, we propose a SplitNN-dedicated data poison attack which does not require to modify any labels but just replaces the local embeddings of a very small number of target-class samples with a carefully constructed trigger vector during training. The experiments on four datasets show that our attack can achieve an attack rate as high as 94%, while bringing negligible side-effects to the model accuracy. Moreover, it is stealthy enough to resist various anomaly detection methods.

Federated learning (FL) is a promising approach for participants’ collaborative learning tasks with cross-silo data. Participants benefit from FL since heterogeneous data can contribute to the generalization of the global model while keeping private data locally. However, practical issues of FL, such as security and fairness, keep emerging, impeding its further development. One of the most threatening security issues is the poisoning attack, corrupting the global model by an adversary’s will. Recent studies have demonstrated that elaborate model poisoning attacks can breach the existing Byzantine-robust FL solutions. Although various defenses have been proposed to mitigate poisoning attacks, participants will sacrifice learning performance and fairness due to strict regulations. Considering that the importance of fairness is no less than security, it is crucial to explore alternative solutions that can secure FL while ensuring both robustness and fairness. This paper introduces a robust and fair model aggregation solution, Romoa-AFL, for cross-silo FL in an agnostic data setting. Unlike a previous study named Romoa and other similarity-based solutions, Romoa-AFL ensures robustness against poisoning attacks and learning fairness in agnostic FL, which has no assumptions of participants’ data distributions and the server’s auxiliary dataset.

Intelligent Applications (iApps), equipped with in-App deep learning (DL) models, are emerging to provide reliable DL inference services. However, in-App DL models are typically compiled into inference-only versions to enhance system performance, thereby impeding the evaluation of DL models. Specifically, the assessment of in-App models currently relies on black-box testing methods rather than direct white-box testing approaches. In this work, we propose TIM, an automated tool designed for conducting large-scale white-box testing of in-App models. Taking an iApp as input, TIM can lift the black-box (i.e., inference-only) in-App DL model into a backpropagation-enabled one and package it together, allowing comprehensive DL model testing or security issues detection. TIM proposes two reconstruction techniques to convert the inference-only model to a backpropagation-enabled version and reconstruct the DL-related IO processing code. In our experiments, we utilize TIM to extract 100 unique commercial in-App models and convert the models to white-box models, enabling backpropagation functionality. Experimental results show that TIM’s reconstruction techniques exhibit high accuracy. We open-source our prototype and part of the experimental data on the websitehttps://zenodo.org/record/7548141.

Adversarial examples that can fool neural network classifiers have attracted much attention. Existing approaches to detect adversarial examples leverage a supervised scheme in generating attacks (either targeted or non-targeted) for training the detectors, which means the detectors are geared to the attacks chosen at the training time and could be circumvented if the adversary does not act as expected. In this paper, we borrow ideas from cryptography and present a novel approach called pseudorandom classifier. In a nutshell, a pseudorandom classifier is a classifier equipped with a mapping to encode the category labels into random multi-bit labels, and a keyed pseudorandom injective function to transform the input to the classifier. The multi-bit labels enable attack-independent and probabilistic detection if the input sample is adversarial. The pseudorandom injection makes the existing white-box adversarial example generation methods, largely based on back-propagation, no longer applicable. We empirically evaluate our method on MNIST, CIFAR10, Imagenette, CIFAR100, and GTSRB. The results suggest that its performance against adversarial examples is comparable to the state-of-the-art.

In this article, we focus on the emerging field of web-based object detection, which has gained considerable attention due to its ability to utilize large amounts of web data for training, thus eliminating the need for labor-intensive manual annotations. However, the noisy and ever-evolving nature of web data poses challenges in preparing high-quality datasets for web-based object detection. To address these challenges, we propose a fully automatic dataset preparation method in this article. Our proposed method incorporates a hierarchical clustering module that assigns multiple precise labels to each image. This module is based on our observation that web image data exhibits different distributions at varying granularities. Furthermore, an evolutionary relabeling module ensures the adaptability of both the prepared dataset and trained detection models to the ever-evolving web data. Extensive experiments demonstrate that our method outperforms other web-based methods, and achieves a comparable performance to those manually labeled benchmark datasets.

The widespread deployment of Deep Learning Inference Services (DLISes) has raised people’s concerns about their data privacy being breached. Although data privacy enhancement has recently attracted a lot of attention, existing solutions all require the cooperation of service providers. Users lose control of their data when making data privacy enhancement decisions. However, it is difficult to enable the user-side control of data abuse prevention because users do not have any programming skills, deep learning knowledge, or rich computing resources. In this work, we propose a fully-automatic userside data privacy enhancement solution, GAPter, for DLISes. Given such a DLIS, GAPter can adaptively fuzz the service for a suitable enhancement strategy, with no cooperation between the DLIS provider and the user. We have implemented and comprehensively evaluated GAPter. The experimental results show that GAPter can find good balance points between privacy enhancement and user data utility.

Various data analysis tasks take frequent item mining as an essential part of them. Thus, it is crucial to design practical privacy-preserving frequent item mining methods such that private data analysis can be enabled. For two-party data analysis tasks, each party possesses its portion of the data and is reluctant to share the data with the other. Although secure computation can enable two-party frequent item mining, the output of exact top-$k$items may still leave the adversary a chance to infer the sensitive information. Differential privacy has been utilized in various data analysis tasks to protect participating individuals, but how to provide differential privacy for two-party frequent item mining has not been addressed. To prevent each party's data from being leaked to the other while achieving differential privacy for releasing the output, we study the problem of differentially private two-party frequent item mining in this paper. We have proposed a practical protocol that can efficiently select the frequent items of the union of two confidential databases in a differentially private way but does not need to combine all the elements. We have conducted extensive experiments to evaluate the proposed method on real-world datasets, and the results show that it can achieve satisfactory utility with affordable overheads.

The phishing scams pose a serious threat to the ecosystem of Ethereum which is one of the largest blockchains in the world. Such a type of cyberattack recently has caused losses of millions of dollars. In this paper, we propose a Self-supervised IncrEmental deep Graph lEarning (SIEGE) model, for the phishing scam detection problem on Ethereum. To overcome the data scalability challenge, we propose splitting the original Ethereum transaction data and constructing transaction graphs for each split. Confronted with the minimal labeled data available, we resort to graph-based self-supervised learning. We design a spatial pretext task to learn high-quality node embeddings inside a single graph split, as well as an incremental learning paradigm and a temporal pretext task to facilitate information flow between different graph splits. To evaluate the effectiveness of SIEGE, we gather a real-world dataset consisting of six-month Ethereum transaction records. The results demonstrate that our model consistently outperforms baseline approaches in both transductive and inductive settings.

Automatic dataset preparation can help users avoid labor-intensive and costly manual data annotations. The difficulty in preparing a high-quality dataset for object detection involves three key aspects: relevance, naturality, and balance, which are not addressed by existing works. In this paper, we leverage information from the web, and propose a fully-automatic dataset preparation mechanism without any human annotation, which can automatically prepare a high-quality training dataset for the detection task with English text terms describing target objects. It contains three key designs, i.e., keyword expansion, data de-noising, and data balancing. Our experiments demonstrate that the object detectors trained with auto-prepared data are comparable to those trained with benchmark datasets and outperform other baselines. We also demonstrate the effectiveness of our approach in several more challenging real-world object categories that are not included in the benchmark datasets.

Payment channels emerge as a promising solution to the scalability issues of blockchain-based digital currency systems, but they implicitly assume that channel owners can periodically monitor the blockchain, which may be impractical for most ordinary users. To address this issue, watchtowers are developed to monitor the blockchain on behalf of its hirers, allowing them to stay offline without security concerns. Despite their usefulness, current watchtower implementations face a security and scalability dilemma. They either require hirers to trust watchtowers, making hirers' funds vulnerable if the watchtower colludes with the counterparty, or require the watchtower to deposit collateral for each hirer, restricting the service scalability due to the watchtower's limited funds. To overcome this dilemma, we propose SilenTower, a mul-tiparty watchtower protocol. It allows a given proportion of collusive protocol participants and provides fund security without collateral. Thus, SilenTower achieves security and scalability simultaneously. Moreover, we propose a quantified definition for watchtower robustness and prove that SilenTower has better robustness than state-of-the-art implementations. We also assess SilenTower's performance through thorough benchmarking and demonstrate that it has lightweight overheads for both partici-pants and hirers.

Butterfly (a cyclic graph motif) counting is a fundamental task with many applications in graph analysis, which aims at computing the number of butterflies in a large graph. With the rapid growth of graph data, it is more and more challenging to do butterfly counting due to the super-linear time complexity and large memory consumption. In this paper, we study I/O-efficient algorithms for doing butterfly counting on hierarchical memory. Existing algorithms of the kind cannot guarantee I/O optimality. Observing that in order to count butterflies, it suffices to "witness" a subgraph instead of the whole structure, a new class of algorithms called semi-witnessing algorithm is proposed. We prove that a semi-witnessing algorithm is not restricted by the lower bound Ømega(|E|2/MB) of a witnessing algorithm, and give a new bound of Ømega(min(|E|2/MB, |E|/|V| √M B)). We further develop the IOBufs algorithm that manages to approach the I/O lower bound, and thus claim its optimality. Finally, we make efforts to parallelize IOBufs to further improve the performance and scalability. We show in the experiment that IOBufs significantly outperforms the state-of-the-art algorithms EMRC and BFC-EM. In addition, IOBufs can scale to conducting butterfly counting on the Clueweb graph with 37 billion edges and quintillions (10^18 ) of butterflies.

Location-based services have significantly affected mobile users’ everyday life, and location privacy has become essential. Some applications (e.g., location-based recommendation, mobility analytics) do not need the raw location data, and the service providers adopt aggregation to protect users’ location traces. However, some works show that even these aggregation data may disclose users’ location privacy when additional prior knowledge is available to an adversary. We consider the location privacy problem in the presence ofLocation Uniqueness, a property by which some geographical locations can be re-identified based on the aggregated point-of-interest information. We first study whether existing protection mechanisms are adequate for defending against this type of attack. Then we present two practical attacks for inferring users’ actual locations based on the POI aggregates. A secure POI aggregate release mechanism is proposed for defending against this type of re-identification attack and achieving differential privacy at the same time. We conduct extensive experiments on real-world datasets. The results show that the existing protection mechanisms cannot provide sufficient protection against location re-identification attacks. The proposed attacks can significantly improve the inference performance, and the proposed protection mechanism achieves satisfactory performance.

ARM TrustZone is widely deployed on commercial-off-the-shelf mobile devices for secure execution. However, many Apps cannot enjoy this feature because it brings many constraints to App developers. Previous works have been proposed to build a secure execution environment for developers on top of TrustZone. Unfortunately, these works are still not a fully-fledged solution for mobile Apps, especially for the emerging intelligent Apps. To this end, we propose LEAP, which is a lightweight developer-friendly TEE solution for mobile Apps. LEAP enables isolated codes to execute in parallel and access peripheral (e.g., mobile GPUs) with ease, flexibly manages system resources upon different workloads, and offers the auto DevOps tool to help developers prepare the codes running on it. We implement the LEAP prototype on the off-the-shelf ARM platform and conduct extensive experiments on it. The experimental results show that Apps can be adapted to run with LEAP easily and efficiently. Compared to the state-of-the-art work along this research line, LEAP can achieve an average 3.57× speedup in supporting intelligent Apps using mobile GPU acceleration.

Mobile edge computing (MEC) is proposed as an extension of cloud computing in the scenarios where the end devices desire better services in terms of response time. Because the edges are usually owned by individuals or small organizations with limited operation capabilities, the data on the edges are easily corrupted (due to external attacks or internal hardware failures). Therefore, it is essential to verify data integrity in the MEC. We propose two Integrity Checking protocols for the mobile Edge storage, called ICE-basic and ICE-batch. Our protocols allow a third-party verifier to check the data integrity on the edges without violating users data privacy and query pattern privacy. We rigorously prove the security and privacy guarantees of the protocols. In addition, we have investigated how to let the end devices cache some verification tags such that the communication cost between end devices and the cloud can be further reduced when a user connects to multiple edges in sequence. We have implemented a proof-of-concept system that runs ICE, and extensive experiments are conducted to evaluate the performance of the proposed protocols. The theoretical analysis and experimental results demonstrate the proposed protocols are efficient both in computation and communication.

Recently, video summarization (VS) techniques are widely used to alleviate huge processing pressure brought by numerous long videos. However, it is hard to summarize long videos efficiently since processing hundreds of frames is still time-consuming. In this paper, we find that the Kernel Temporal Segmentation (KTS) method designed for detecting the shot boundaries in SOTA VS methods is time-consuming while handling long videos. To address this issue, we propose the Distribution-based KTS (D-KTS) by fully considering the characteristic of shot length distribution. Furthermore, we propose the Hash-based Adaptive Frame Selection (HAFS) to improve the system performance by fully taking advantage of the temporal locality of long videos. Our experiments present that the proposed D-KTS is 92.70% faster and takes up 90.08% less memory than the baseline KTS method on average.

Many mobile apps access users’ trajectories to provide critical services (e.g., trip tracking). Unfortunately, in such apps, malicious users may upload fake trajectories to cheat providers for illegal benefits. There are few works in the literature that delicately study trajectory forgery problems. In this paper, we first take the perspective of attackers and consider how they fabricate vivid trajectories confronting a strict provider. In particular, we use the technique of adversarial examples in deep learning to propose a trajectory forgery method, which produces fake trajectories satisfying two conditions: (1) having the motion characteristics indistinguishable from those of real ones, and (2) matching a reasonable walking, cycling, or driving route when being projected to the map. We show through experiments that they can hardly be detected by mainstream trajectory service providers, even after being equipped with machine learning-based approaches. Therefore, we further present a dedicated countermeasure by validating the reasonability of reported received signal strength indicator (RSSI) data of WiFi access points (APs) nearby every location. It can well deal with the most challenging replay scenario, which can hardly be handled by existing WiFi-based location verification methods. We conduct extensive real-world experiments in three local commercial areas covering walking, cycling, and driving scenarios. Results demonstrate the high detection accuracy of this method.

Redundancy elimination (RE) systems allow network users to remove duplicate parts in their messages by introducing caches at both message senders’ and receivers’ sides. While RE systems have been successfully deployed for handling unencrypted traffic, making them work over encrypted links is still open. A few solutions have been proposed recently, however they either completely violate end-to-end security or focus on single-user setting. In this paper, we present a highly secure RE solution which supports cross-user redundancy eliminations on encrypted traffics. Our solution not only preserves the end-to-end security against outside adversaries, but also protects users’ privacy against semi-honest RE agents. Furthermore, our solution can defend malicious users’ poisoning attack, which is crucial for cross-user RE systems but has never been studied before. In cross-user RE systems, since all users inside a LAN write into a shared, global cache and use it to recover their original messages from deduplicated ones, the poisoning attack is prone to happen, and cause systematic damage to all users even when only one user is malicious and injects poisoned data into the cache. We rigorously prove our solution’s security properties, and demonstrate its promising performance via testing the proof-of-concept implementation with real-world internet traffic data.

Federated learning, in contrast to traditional learning paradigms, has demonstrated its unique advantages in providing intelligence at the edge. However, existing federated learning approaches focus on the end-to-end classification tasks requiring a simple collaboration procedure where each participant can perform its local training independently. Unfortunately, there are still many tasks relying on learning the distinguishable feature metrics with respect to all the data, which is a different collaboration procedure across training participants. For example, the model for people identification has to ensure the feature representing a person is dissimilar to those representing others. To enable such federated learning for deep metrics (a.k.a federated deep metric learning) is challenging due to the data privacy and procedure robustness issues. With the consideration of these two challenges, this work proposes a novel computing framework for federated deep metric learning. This framework leverages the system-algorithm co-design to address privacy concerns via the Trusted Execution Environment (SGX enclave) and Differential Privacy mechanism. It also introduces a large-scale federated protocol which can robustly and efficiently deal with practical factors like the network fluctuation. We implement and evaluate our computing framework with two settings. One is a real-world implementation with a large number of mobile devices, while the other one is in our controllable environment for conducting experiments in various tasks. Our evaluation results show that our computing framework is able to train federated deep metric learning models with excellent scalability, data privacy preserving, and considerable accuracy even in exception conditions.

In software defined networking, theflatdesign of distributed control plane enables the management of multi-domain networks that are incapable of deploying a root controller. However, it is very difficult to avoid policy conflicts between independent local controllers due to the lack of centralized arbitration. Moreover, domains without trust may not be always cooperative and could even cheat to maximize their own interests. In this article, we first consider the cooperative scenario and address the problem of traffic engineering in a flat distributed control plane. We propose a fully distributed algorithm, calledDisTE, which can provide max-min fair bandwidth allocation for flows and maximize resource utilization.DisTEalso preserves the local topology of each domain and achieves policy consistency by multiple rounds of synchronization. We then consider the non-cooperative scenario, where selfish domains may discriminate bandwidth requests from other domains or overstate theirs owns to squeeze more bandwidths.

As the mobile era matures, it is increasingly competitive to market mobile apps, forcing companies to invest heavily on mobile user acquisition campaigns. This has unfortunately given birth to a new form of Internet fraud, which we refer to as “app distribution fraud”. This new fraud involves collusion between ISPs and fraudulent app distributors where app download is hijacked/redirected. In this article, we have the unique opportunity to cooperate with a major e-commerce company (with about 0.2 billion active users per month) to take a first peek at this problem. Through the nationwide measurement results, we find that app distribution fraud is ubiquitous yet stealthy — about 1.55 percent app downloads are hijacked/redirected, affecting more than 75 percent of the cities we tested and causing an estimated 7.46 billion U.S. dollars financial loss per year. We follow up with additional measurements on the technical mechanism of the fraud and the scope of the fraud (i.e., what other apps are also affected). Surprisingly, we find that sometimes the original app a user intends to download can be replaced with a completely different app, rendering the user's device at risks.

Android is a growing target for malicious software (malware) because of its popularity and functionality. Malware poses a serious threat to users’ privacy, money, equipment and file integrity. A series of data-driven malware detection methods were proposed. However, there exist two key challenges for these methods: (1) how to learn effective feature representation from raw data; (2) how to reduce the dependence on the prior knowledge or human labors in feature learning. Inspired by the success of deep learning methods in the feature representation learning community, we propose a malware detection framework which starts with learning rich-features by a novel unsupervised feature learning algorithm Merged Sparse Auto-Encoder (MSAE). In order to extract more compact and discriminative feature from the rich-features to further boost the malware detection capability, a hybrid deep network learning algorithm Stacked Hybrid Learning MSAE and SDAE (SHLMD) is established by further incorporating a classical deep learning method Stacked Denoising Auto-encoders (SDAE). After that, we feed the feature learned by MSAE and SHLMD respectively to classification algorithms, e.g., Support Vector Machine (SVM) or K-NearestNeighbor (KNN), to train a malware detection model. Evaluation results on two real-world datasets demonstrate that SHLMD achieves 94.46 and 90.57 percent accuracy respectively, which outperforms the classical unsupervised feature representation learning Sparse Auto-encoder (SAE). MSAE performs similarly to SAE. SHLMD can further improve the performance of MSAE and the supervised fine-tuned method SDAE. Besides, we compare the performance of our methods with that of state-of-the-art detection approaches, including classical deep-learning-based methods. Extensive experiments show that our proposed methods are effective enough to detect Android malware.

Vast data and computing resources are commonly needed to train deep neural networks, causing an unaffordable price for individual users. Motivated by the increasing demands of deep learning applications, sharing well-trained models becomes popular. The owner of a pre-trained model can share it by publishing the model directly or providing a prediction interface. Either way, individual users can benefit from deep learning without much cost, and computing resources can be saved. However, recent studies of machine learning security have identified severe threats to these model publishing approaches. This paper will focus on the privacy leakage issue of publishing well-trained deep neural network models. To tackle this problem, we propose a series of secure model publishing solutions based on training task parallelism. Specifically, we show how to estimate private model parameters through parallel model training and generate new model parameters in a privacy-preserving manner to replace the original ones for publishing. Based on data parallelism and parameter generating techniques, we design another two solutions concentrating on model quality and parameter privacy, respectively. Through privacy leakage analysis and experimental attack evaluation, we conclude that deep neural network models published with our solutions can provide on-demand model quality guarantees and resist membership inference attacks.

Location-based services have significantly affected mobile users' everyday life, and location privacy is also an essential issue in these services. In some applications (e.g., location-based recommendation, mobility analytic), the raw data is not required, and the service providers adopt aggregation to protect users' location traces. However, some works show that even these aggregation data may disclose users' location privacy when other prior knowledge is available to an adversary. We consider the location privacy problem in the presence of Location Uniqueness, which is a property that some geographical locations can be re-identified based on the aggregated point-of-interest (POI) information. We first study whether previous protection mechanisms are effective for defending against this novel type of attack. Then we present two practical attacks for inferring users' actual locations based on the POI aggregates. Furthermore, we propose a secure POI aggregate release mechanism that can defend against this type of re-identification attack and achieve differential privacy at the same time. We conduct extensive experiments on real-world datasets. The results show that the existing protection mechanisms cannot provide sufficient protection. The proposed enhanced attacks can significantly improve the inference performance, and the proposed protection mechanism achieves satisfactory performance.

Payment Channel Network (PCN) is one of the most promising technologies for scaling the capacity of blockchain-based cryptocurrencies and improving the quality of blockchain-based services. However, during the use of PCNs, a significant portion of the payment channels gradually become exhausted, which triggers additional consumption of on-chain resources and makes PCNs less useful. This is a fundamental problem for blockchain-based cryptocurrencies, worthy of a thorough investigation.In this paper, we propose OPRE, a protocol for OPtimal off-chain REcovering of payment channels, to solve this problem. It is optimal in that it recovers the maximum number of nearly exhausted channels in the PCN. Furthermore, we consider users’ privacy concerns and design a privacy-preserving version of this protocol, so that users’ balance information does not need to be revealed. This protocol maintains optimality in recovering payment channels while providing cryptographically strong privacy guarantee. In addition to the theoretical design and analysis, we also implement OPRE and experimentally evaluate its performance. The results show that the OPRE protocol is both efficient and effective.

As Video Streaming and Analytics (VSA) systems become increasingly popular, serious privacy concerns have risen on exposing too much unnecessary private information to the VSA providers. Yet, it is challenging to protect privacy while still preserving desired VSA features, i.e., effective analytics, forensic support, resource efficiency, and real-time execution. In this paper, we present a VSA privacy enhancement system (PECAM), which addresses the above challenge with no change in the VSA back-end. PECAM leverages a novel Generative Adversarial Network to perform the privacy-enhanced securely-reversible video transformation. PECAM also incorporates a couple of system optimizations into its VSA workflow to reduce network bandwidth usage and enable real-time processing on cameras. We implement our PECAM prototype on commodity hardware and evaluate its performance via both security study and extensive experiments. Results demonstrate that PECAM can effectively enhance the visual privacy of VSA in the presence of an adversary, and its transformed videos, when taken as input for various VSA back-end tasks, maintain a 96% accuracy of corresponding original videos. Additionally, it performs 12.3× and 1.8× better than baseline methods in terms of the computing cost and network bandwidth usage, respectively.

Android shared storage is shared with all the applications (apps for short) and the user. It is common to see that a large amount of apps store different kinds of files on it. It is well known that apps granted the read or write permissions can freely access any files in the shared storage. As a consequence, the shared storage has been demonstrated to expose sensitive information and jeopardize users' privacy. In this paper, we systematically study a simple but overlooked threat related to the shared storage-the lack of input validation (e.g., integrity verifications) when consuming files on the shared storage. We argue that the untrusted input from the shared storage is a much ubiquitous problem. By undertaking an empirically study through a static analysis tool we develop, we find over 30 percent of the 13,746 analyzed popular apps on the market suffer from such problem. By investigating the types of files consumed, we find shockingly a large fraction of apps store and consume sensitive files, which allows us to construct end-to-end attacks. Considering the ubiquity of this class of vulnerabilities, we finally define better access control policies for external storage to eliminate them for most apps.

As one of a few critical technologies to cloud storage service, deduplication allows cloud servers to save storage space by deleting redundant file copies. However, it often leaks side channel information regarding whether an uploading file gets deduplicated or not. Exploiting this information, adversaries can easily launch a template side-channel attack and severely harm cloud users' privacy. To thwart this kind of attack, we resort to the k-anonymity privacy concept to design secure threshold deduplication protocols. Specifically, we have devised a novel cryptographic primitive called “dispersed convergent encryption” (DCE) scheme, and proposed two different constructions of it. With these DCE schemes, we successfully construct secure threshold deduplication protocols that do not rely on any trusted third party. Our protocols not only support confidentiality protections and ownership verifications, but also enjoy formal security guarantee against template side-channel attacks even when the cloud server could be a “covert adversary” who may violate the predefined threshold and perform deduplication covertly. Experimental evaluations show our protocols enjoy very good performance in practice.

Link Layer Discovery Protocol (LLDP), which is widely used by the controller in Software-Defined Networking to discover the network topology, has been demonstrated to be unable to guarantee the integrity of its messages. Attackers could exploit this vulnerability to fabricate LLDP packets to declare a false link connecting two distant switches to the controller. By doing so, the controller would be misled to route flows to the false links, which leads to further DoS, eavesdropping and even hijacking attacks. This attack seems very similar to the well-known Worm-Hole Attack in wireless sensor networking (WSN). Nevertheless, in WSN, attackers are assumed to leverage an out-of-band wired channel to achieve the true packet transmission between the two cheating sensor nodes. Unfortunately, in SDN, there usually does not exist any out-of-band channels between the distant cheating switches. Flows misguided to the fake link will cause 100% packet loss, and thus be detected soon. In this article, we address this problem and propose the first True worm-hole attack in SDN, which could achieve packet transmission over the forged link without using any out-of-band channels. Instead, it introduces a relay host in the networks to build a completely in-band covert channel between the two cheating switches. Unlike the existing studies, a relay host is not required to be directly linked to them. Moreover, attackers are only assumed to poss the remote read and write privileges of the flow tables of the both cheating switches and do not have to alter any of their software or hardware. Our extensive experiments demonstrate the high feasibility of this attack. Both the increases of transmission delays and packet loss rates are within a reasonable range. We finally present and evaluate the countermeasures against the proposed attack.

Deep neural networks (DNNs) have brought significant performance improvements to various real-life applications. However, a DNN training task commonly requires intensive computing resources and a huge data collection, which makes it hard for personal devices to carry out the entire training, especially for mobile devices. The federated learning concept has eased this situation. However, it is still an open problem for individuals to train their own DNN models at an affordable price. In this article, we propose an alternative DNN training strategy for resource-limited users. With the help of an untrusted server, end users can offload their DNN training tasks to the server in a privacy-preserving manner. To this end, we study the possibility of the separation of a DNN. Then we design a differentially private activation algorithm for end users to ensure the privacy of the offloading after model separation. Furthermore, to meet the rising demand for federated learning, we extend the offloading solution to parallel DNN models training with a secure model weights aggregation scheme for the privacy concern. Experimental results prove the feasibility of computation offloading solutions for DNN models in both solo and parallel modes.

Triangle counting is a building block for a wide range of graph applications. Traditional wisdom suggests that i) hashing is not suitable for triangle counting, ii) edge-centric triangle counting beats vertex-centric design, and iii) communication-free and workload balanced graph partitioning is a grand challenge for triangle counting. On the contrary, we advocate that i) hashing can help the key operations for scalable triangle counting on Graphics Processing Units (GPUs), i.e., list intersection and graph partitioning, ii) vertex-centric design reduces both hash table construction cost and memory consumption, which is limited on GPUs. In addition, iii) we exploit graph and workload collaborative, and hashing-based 2D partitioning to scale vertex-centric triangle counting over 1000 GPUs with sustained scalability. In this article, we present Trust which performs triangle counting with the hash operation and vertex-centric mechanism at the core. To the best of our knowledge, Trust is the first work that achieves over one trillion Traversed Edges Per Second (TEPS) rate for triangle counting.

Distributed learning unleashes the power of training collaboration among multiple parties who have different training data. While participants enjoy mutually beneficial outcomes of the distributed learning, which cannot be achieved by single party, they also worry about the risk of privacy leaking. In fact, recent work shows that a malicious participant is able to leverage a Generative Adversarial Network (GAN) to steal sensitive information of training data owned by others through shared gradient updates. However, existed countermeasures, such as the differential privacy or cryptographic methods, could disturb the training in terms of model accuracy or computation overhead. In this paper, we seek to mitigate this privacy issue in a non-intrusive manner. Instead of passive protection, we propose to actively detect such GAN-based attackers at the very beginning of training. Our detection only utilizes the gradient updates uploaded by participants during the training, so it is transparent to participants and does not require protocol changes. We demonstrate the effectiveness of our detection through extensive experiments in different settings and attack scenarios.

In recent years, more and more mobile applications adopt deep learning technologies, especially CNN-based image recognition. To protect service providers' interests, the CNN models are usually deployed on the cloud, and the users are required to upload raw images, which cause serious privacy concerns since images may contain sensitive information unrelated to the desired recognition tasks. The previous solution off-loads the shallow portions of the CNN to the clients, and thus the uploaded data becomes the extracted lower-level features rather than the raw images. Nevertheless, although service providers are prevented from obtaining the original images, it is still probably for them to perform some sensitive recognition tasks other than the desired one on the lower-lever features (even after being perturbed to satisfy Differential Privacy). Different from such solution, in this paper, we propose an independent local CNN, which is dedicated for the image perturbation on the clients. It is co-trained with the cloud CNN to learn to intelligently allocate diverse noises among pixels depending on their significance to the desired recognition service. Extensive experiments demonstrate that our mechanism can well prevent curious service providers from performing undesired recognition tasks while maintaining the high accuracy of the desired one.

In order to perform deep learning tasks everywhere, many optimizations have been proposed to address the resource limitations on mobile systems like IoTs. A key approach among others is to dynamically adjust computational resources of the deep learning inference according to the characteristics of incoming inputs. For example, one of popular optimizations is to pick for each input a suitable combination of computations with respect to its inference difficulty. However, we find out that such “dynamic routing” of computations could be exploited to drain/waste precious resources on mobile deep learning systems. In this work, we introduce a new deep learning attack dimension, the computational resources draining, and demonstrate its feasibility in one of possible attack manners, the adversarial examples of input data. We describe how to construct our special adversarial examples aiming to the resource draining, and show that these poisoned inputs are able to increase the computation loads on purpose with two experiment datasets. We hope that our findings can shed light on the path of improving the robustness of mobile deep learning optimizations.

Machine learning as a service has emerged recently to relieve tensions between heavy deep learning tasks and increasing application demands. A deep learning service provider could help its clients to benefit from deep learning techniques at an affordable price instead of huge resource consumption. However, the service provider may have serious concerns about model privacy when a deep neural network model is published. Previous model publishing solutions mainly depend on additional artificial noise. By adding elaborated noises to parameters or gradients during the training phase, strong privacy guarantees like differential privacy could be achieved. However, this kind of approach cannot give guarantees on some other aspects, such as the quality of the disturbingly trained model and the convergence of the modified learning algorithm. In this paper, we propose an alternative private deep neural network model publishing solution, which caused no interference in the original training phase. We provide privacy, convergence and quality guarantees for the published model at the same time. Furthermore, our solution can achieve a smaller privacy budget when compared with artificial noise based training solutions proposed in previous works. Specifically, our solution gives an acceptable test accuracy with privacy budget ϵ = 1. Meanwhile, membership inference attack accuracy will be deceased from nearly 90% to around 60% across all classes.

Real-time user emotion recognition is highly desirable for many applications on eyewear devices like smart glasses. However, it is very challenging to enable this capability on such devices due to tightly constrained image contents (only eye-area images available from the on-device eye-tracking camera) and computing resources of the embedded system. In this paper, we propose and develop a novel system called EMO that can recognize, on top of a resource-limited eyewear device, real-time emotions of the user who wears it. Unlike most existing solutions that require whole-face images to recognize emotions, EMO only utilizes the single-eye-area images captured by the eye-tracking camera of the eyewear. To achieve this, we design a customized deep-learning network to effectively extract emotional features from input single-eye images and a personalized feature classifier to accurately identify a user's emotions. EMO also exploits the temporal locality and feature similarity among consecutive video frames of the eye-tracking camera to further reduce the recognition latency and system resource usage. We implement EMO on two hardware platforms and conduct comprehensive experimental evaluations. Our results demonstrate that EMO can continuously recognize seven-type emotions at 12.8 frames per second with a mean accuracy of 72.2%, significantly outperforming the state-of-the-art approach, and consume much fewer system resources.

Debugging a traditional network is notoriously difficult due to network devices' heterogeneity and protocols' decentralized nature, but Software-Defined Networking (SDN) is changing this predicament. Recent works have provided very nice approaches for an administrator to perform several fundamental network tests in a single-domain SDN network. However, how to perform these tests securely in multi-domain networks still remains open. In this paper, we study the highly challenging problem of inter-domain forwarding loop test in a SDN environment. We present two novel testing protocols that can be used for inter-domain loop tests. Both protocols are secure in the sense that they protect each domain's private information about its topology and configuration. The first protocol, based on random sampling, is highly efficient with a small error probability diminishing exponentially in the sample size. The second protocol, based on secure set intersection test, guarantees 100 percent accuracy of the result, although not as efficient as the first one. We provide rigorous proofs for the security and accuracy guarantees, and show our protocols have very good efficiency by testing them with real-world network data.

Face recognition has become increasingly popular in recent years. However, in some special cases, many face recognition calculations cannot be performed effectively due to the lack of sufficient computing power of the terminal, which poses a challenge to the practical application of face recognition technology. Cloud computing provides a good platform for solving this problem due to its abundant computing resources. However, cloud computing poses new challenges, such as how to protect clients' data privacy without reducing efficiency. In this paper, we review some of the results of previous research and analyze an outsourcing protocol for eigen decomposition and singular value decomposition. On this basis, we propose a secure and efficient outsourcing protocol for face recognition through principal component analysis. In the proposed protocol, information privacy is well protected, and computational resources are saved by means of conversions of the original image information. In addition, local verification is supported to cope with the laziness of the cloud. We show the feasibility and advancement of our protocol from both theoretical and experimental perspectives.

Multi-User MIMO (MU-MIMO) has attracted much attention due to its significant advantage of increasing the utilization ratio of wireless channels. However, Frequency-Division Duplex (FDD) systems are vulnerable to eavesdropping, since the explicit CSI feedback can be manipulated. In this paper, we show that Time-Division Duplex (TDD) systems are insecure as well. In particular, we show that it is possible to eavesdrop on other users' downloads by tuning training sequences. In order to defend MU-MIMO against such threats, we propose a secure CSI estimation scheme, which can provide correct estimates of CSI when adversarial users are in presence. We prove that our scheme is secure against training sequence based eavesdropping attack. We have implemented our scheme for TDD MU-MIMO systems and performed a series of experiments. Results demonstrate that our secure CSI estimation scheme is highly effective in protecting TDD MIMO networks against eavesdropping attack. Furthermore, we extend our scheme to support massive MU-MIMO networks, with a carefully redesigned uplink protocol and optimized power allocation to achieve higher spectral efficiency. To be more practical, we also take mismatch channel issue into our consideration. An enhancement scheme is proposed and we show that our scheme with enhancement is secure and correct under mismatch channel.

Mobile edge computing (MEC) is proposed as an extension of cloud computing in the scenarios where the end devices desire better services in terms of response time. Edge nodes are deployed at the proximity of the end devices, and it can pre-download parts of data stored in the cloud so that the end devices can access these data with low latency. However, because the edges are usually owned by individuals and small organizations, which have limited operation capacities for maintaining the machines, the data on the edges are easily corrupted (due to external attacks or internal hardware failures). Therefore, it is essential to verify data integrity in the MEC. We propose two Integrity Checking protocols for mobile Edge computing, called ICE-basic and ICE-batch, which are designed for the cases where the user wants to check data integrity on a single edge or multiple edges, respectively. Based on the concept of provable data possession and the technique of private information retrieval, our protocols allow a third-party verifier to check the data integrity on the edges without violating users' data privacy and query pattern privacy. We rigorously prove the security and privacy guarantees of the protocols. Furthermore, we have implemented a proof-of-concept system that runs ICE, and extensive experiments are conducted. The theoretical analysis and experimental results demonstrate the proposed protocols are efficient both in computation and communication.

The increasing scale of software defined networks (SDN) raises the requirement of distributed control plane, for providing scalable, reliable and high performance network management capabilities. In particular, the flat design of distributed control plane enables the management of networks with multiple independent domains that are incapable of deploying a root controller. However, it is very difficult to avoid policy conflicts between multiple controllers in flat plane due to the lack of arbitration. In this paper, we address the problem of traffic engineering in a flat control plane, and design a distributed traffic engineering algorithm, called DisTE, which can provide max-min fair bandwidth allocation for flows and maximize the resource utilization, using a fully distributed arbitration mechanism. DisTE also preserves the local topology of each domain using the topology aggregation method, and supports consistency by multiple rounds of synchronizations. We examine four strategies for determining the synchronization timings, and find that linearly decreasing interval method provides a better trade-off between network utilization and time costs. Experiments on a 717-switches 5-domain network topology demonstrate that DisTE could drive the link utilization ratio to more than 93%, and reduce up to 95% convergence time at cost of 3% relative error on fairness, compared to the centralized approach.

Today's large-scale enterprise networks, data center networks, and wide area networks can be decomposed into multiple administrative or geographical domains. Domains may be owned by different administrative units or organizations. Hence protecting domain information is an important concern. Existing general-purpose Secure Multi-Party Computation (SMPC) methods that preserves privacy for domains are extremely slow for cross-domain routing problems. In this paper we present PYCRO, a cryptographic protocol specifically designed for privacy-preserving cross-domain routing optimization in Software Defined Networking (SDN) environments. PYCRO provides two fundamental routing functions, policy-compliant shortest path computing and bandwidth allocation, while ensuring strong protection for the private information of domains. We rigorously prove the privacy guarantee of our protocol. To improve time efficiency we design the QuIck Pathing (QIP) technique. QIP only requires one-time offline preprocessing and very fast online computation. We have implemented a prototype system that runs PYCRO and QIP on servers in a campus network. Experimental results using real ISP network topologies show that PYCRO and QIP are very efficient in computation and communication costs.

Due to the loose authentication requirement between access points (APs) and clients, it is notoriously known that WLANs face long-standing threats such as rogue APs and network freeloading. Take the rogue AP problem as an example, unfortunately encryption alone does not provide authentication. APs need to be equipped with certificates that are trusted by clients ahead of time. This requires either the presence of PKI for APs or other forms of pre-established trust (e.g., distributing the certificates offline), none of which is widely used. Before any strong security solution is deployed, we still need a practical solution that can mitigate the problem. In this paper, we explore a non-cryptographic solution that is readily deployable today on end hosts (e.g., smartphones and laptops) without requiring any changes to the APs or the network infrastructure. The solution infers the Carrier Frequency Offsets (CFOs) of wireless devices from Channel State Information (CSI) as their hardware fingerprints without any special hardware requirement. CFO is attributed to the oscillator drift, which is a fundamental physical property that cannot be manipulated easily and remains fairly consistent over time but varies significantly across devices. The real experiments on 23 smartphones and 34 APs (with both identical and different brands) in different scenarios demonstrate that the detection rate could exceed 94%.

The great potential of mobile crowdsourcing has started to attract attention of both industries and the research community. However, current commercial mobile crowdsourcing marketplaces are unsatisfactory because of the limited worker base and functionality. In this paper, we first revisit the foundation of performing mobile crowdsourcing on location-based social networks (LBSNs) through specially designed survey studies and comparison experiments involving hundreds of users. Our results reveal that active check-ins are good indicators of picking a right user to perform tasks, and LBSN could be an ideal platform for mobile crowdsourcing given proper services provided. We then propose both the centralized and decentralized design of MobiCrowd, a mobile crowdsourcing service built on LBSNs. Our evaluation, through trace-driven simulation and real-world experiments, demonstrates that the proposed schemes can effectively find workers for mobile crowdsourcing tasks associated with different venues by analyzing their location check-in histories.

As location-based services (LBSs) on smartphones become increasingly popular, such services are causing serious privacy concerns, because many users are unwilling to see their location information leaked to service providers. Recently, in order to protect users’ location privacy, researchers have introducedgeo-indistinguishability, the first specialized privacy model for LBSs that can provide provable privacy guarantees. Intuitively, geo-indistinguishability means that through perturbation, any two locations within a given distance produce observations with similar distributions, and thus, attackers have no way to learn users’ real locations. However, even if geo-indistinguishability is achieved, there remains a significant threat to users’ location privacy: the privacy consumption increases with the number of queries for the existing geo-indistinguishable location perturbation mechanism, and therefore, there is a high risk of privacy violation when the number of queries is not small. In this paper, we enhance the privacy protection for LBSs by proposing an improved geo-indistinguishable mechanism. It can reduce the privacy costs to almost 0 when the user’s location satisfies a condition. We also present an improvement to further reduce the privacy costs when the above condition is not satisfied. Evaluations upon two public trace data sets show that the proposed mechanisms can dramatically save the privacy budget and thus support much more queries. The results also show that the proposed mechanisms are efficient, and their performance is controllable.

With the popularization of human-carried devices, it is possible for Mobile Crowd Sensing (MCS) applications to perform large scale sensing with sensors embedded in these devices. Many incentive mechanisms for MCS applications have been proposed due to the importance of attracting more worker to participate. However, these mechanisms failed to consider situations where there are constraints on both crowdsourcer's budget and workers' capacity. In contrast, we design mechanisms which ensure approximately maximized value of services provided by selected workers under both constraints based on reverse auctions. Not only do we study the scenario where every worker is trusted and will not infer others' bids, but we also investigate the scenario where there are some honest-but-curious workers. For the former, we design a truthful, individual rational, and computationally efficient incentive mechanism that achieves nearly optimal benefits. For the latter, we design an approximately truthful, individual rational, computationally efficient, and differentially private incentive mechanism that helps to protect workers' privacy from the infringement of curious workers and achieves nearly optimal benefits. Rigorous theoretical analyses and extensive simulations are given to validate the above properties and evaluate the performance of our incentive mechanisms.

Protecting the privacy of mobile phone user participants is extremely important for mobile phone sensing applications. In this paper, we study how an aggregator can expeditiously compute the minimum value or the kth minimum value of all users' data without knowing them. We construct two secure protocols using probabilistic coding schemes and a cipher system that allows homomorphic bitwise XOR computations for our problems. Following the standard cryptographic security definition in the semi-honest model, we formally prove our protocols' security. The protocols proposed by us can support time-series data and need not to assume the aggregator is trusted. Moreover, different from existing protocols that are based on secure arithmetic sum computations, our protocols are based on secure bitwise XOR computations, thus are more efficient.

Motion sensors, especially accelerometers, on smartphones have been discovered to be a powerful side channel for spying on users' privacy. In this paper, we reveal a new accelerometer-based side-channel attack which is particularly serious: malware on smartphones can easily exploit the accelerometers to trace metro riders stealthily. We first address the challenge to automatically filter out metro-related data from a mass of miscellaneous accelerometer readings, and then propose a basic attack which leverages an ensemble interval classifier built from supervised learning to infer the riding trajectory of the user. As the supervised learning requires the attacker to collect labeled training data for each station interval, this attack confronts the scalability problem in big cities with a huge metro network. We thus further present an improved attack using semi-supervised learning, which only requires the attacker to collect labeled data for a very small number of distinctive station intervals. We conduct real experiments on a large self-built dataset, which contains more than 120 h of data collected from six metro lines of three major cities. The results show that the inferring accuracy could reach 89% and 94% if the user takes the metro for four and six stations, respectively. We finally discuss possible countermeasures against the proposed attack.

Collaborative spectrum sensing has become increasingly popular in cognitive radio networks to enable unlicensed secondary users to coexist with the licensed primary users and share spectrum without interference. Despite its promise in performance enhancement, collaborative sensing is still facing a lot of security challenges. The problem of revealing secondary users' location information through sensing reports has been reported recently. Unlike any existing work, in this paper we not only address the location privacy issue in the collaborative sensing to be against semi-honest adversaries, but also take malicious adversaries into consideration. We propose efficient schemes to protect secondary users' reports from being revealed in the aggregation process at the fusion center. We rigorously prove that our privacy-preserving collaborative sensing schemes are secure against attacks from both the fusion center and secondary users. We also evaluate our schemes extensively and verify its efficiency and feasibility.

Ridesharing services have gained tremendous popularity in recent years, benefiting the traffic and environment of cities to a large extent. However, with the demand of ridesharing services increasing sharply, serious privacy concerns (e.g., users' mobility patterns) of ridesharing have become a major barrier against its further development. In this paper, we study the privacy protection of users' location information in the scheduling of ridesharing services. Based on a state-of-the-art variant of differential privacy, joint differential privacy, we first propose a scheduling protocol for the purpose of protecting users' location privacy and minimizing vehicle miles in the system. Then, in order to obtain a practical solution, we investigate several techniques to enhance the proposed protocol from both the privacy and efficiency aspects. The privacy of the proposed scheduling protocol is rigorously proven. Furthermore, we extensively evaluate our proposal based on a real-world data set. The analysis and experimental results show that the proposed protocol can achieve joint differential privacy, satisfactory scheduling performance, and reasonable efficiency.

Multi-User MIMO has attracted much attention due to its significant advantage of increasing the utilization ratio of wireless channels. Recently a serious eavesdropping attack, which exploits the CSI feedback of the FDD system, is discovered in MU-MIMO networks. In this paper, we firstly show a similar eavesdropping attack for the TDD system is also possible by proposing a novel, feasible attack approach. Following it, a malicious user can eavesdrop on other users' downloads by transforming training sequences. To prevent this attack, we propose a secure CSI estimation scheme for instantaneous CSI. Furthermore, we extend this scheme to achieve adaptive security when CSI is relatively statistical. We have implemented our scheme for both uplink and downlink of MU-MIMO and performed a series of experiments. Results show that our secure CSI estimation scheme is highly effective in preventing downlink leakage against malicious users.

Device-to-device communication is important to emerging mobile applications such as Internet of Things and mobile social networks. Authentication and key agreement among multiple legitimate devices is the important first step to build a secure communication channel. Existing solutions put the devices into physical proximity and use the common radio environment as a proof of identities and the common secret to agree on a same key. However they experience very slow secret bit generation rate and high errors, requiring several minutes to build a 256-bit key. In this work, we design and implement an authentication and key agreement protocol for mobile devices, called The Dancing Signals (TDS), being extremely fast and error-free. TDS uses channel state information (CSI) as the common secret among legitimate devices. It guarantees that only devices in a close physical proximity can agree on a key and any device outside a certain distance gets nothing about the key. Compared with existing solutions, TDS is very fast and robust, supports group key agreement, and can effectively defend against predictable channel attacks. We implement TDS using commodity off-the-shelf 802.11n devices and evaluate its performance via extensive experiments. Results show that TDS only takes a couple of seconds to make devices agree on a 256-bit secret key with high entropy.

Anonymous content sharing over the Internet protects user privacy and content confidentiality. Most overlay anonymous communication protocols employ some relay nodes as the proxies to forward content and require relays to perform cryptography or coding operations on messages. They have two major limitations. First, extra computation overhead may discourage overlay nodes from serving as relays. Second, long forwarding latency at relays makes an anonymous path easier to fail under network churn. In this paper, we present a lightweight and decentralized anonymous content sharing system named Garlic Cast, which requires near-zero computation cost on relays and does not rely on any centralized service. Garlic Cast uses random walks to find proxies in overlay networks and an security-enhanced Information Dispersal Algorithm to search and deliver content files. We have implemented a prototype of Garlic Cast and performed extensive simulation on real overlay topologies. Evaluation results show that the throughput of Garlic Cast is higher than that of RSA-based anonymous routing by over two orders of magnitude. Garlic Cast provides high level of anonymity and is robust to various attacks.

Due to the asynchronous and distributed nature of the data plane, consistent configuration updating across multiple switches is a challenging issue in Software-Defined Networking (SDN). The existing version-stamping-based mechanism (VSM) could guarantee per-packet consistency, but this mechanism is designed for non-adversarial settings and can be compromised easily by a malicious attacker. In this paper, we propose an efficient flow-ordered update mechanism that aims to provide per-packet consistency in adversarial settings. Our proposal does not need to stamp data packets with the configuration version, and is robust against both the packet-tampering and packet-dropping attacks. It outperforms a naive mechanism that simply patches VSM using digital signatures in three aspects: First, the switches in this mechanism only need to sign and verify a single control packet, which significantly improves the packet processing time. Second, it avoids keeping both old and new policies on switches during the update, and thus achieves better space efficiency. Third, it reduces the time delay for new policies to come into force. We evaluate our mechanism on a self-constructed SDN testbed and the results demonstrate high efficiency.

Offloading part of cellular traffic through existing alternative wireless networks, such as femtocells and WiFi networks, is one promising solution to the severe traffic overload faced by cellular network providers (CSPs) nowadays. Most existing cellular offloading auction mechanisms assume the CSP has the knowledge of incoming overloaded traffic demand, and satisfy the demand by offloading. However, in practice, with the explosive growth of mobile device communications, the overloaded traffic demand at CSPs is very likely to pass over the total capability that third-party resource owners can provide. Then it is critical to enable CSPs to optimize the traffic handling capacity gain through offloading with budget constraints. In this paper, we propose two efficient Competitive Auction MEchanisms for mObile offloading, CAMEO-min and CAMEO-ws. Both mechanisms are proven to be non-budget-deficit, individually rational and incentive-compatible, and have guaranteed lower bounds on the ratio of the CSP's gain achieved in them to the maximum gain that the CSP could achieve in any omniscient auction (the auction with an omniscient auctioneer). Our extensive evaluations show that CAMEOs achieve very good performance in terms of the maximization of the CSP's gain especially when the global bidder dominance or the region dominance is big.

Being an important application of spectrum sharing in cellular networks, mobile traffic offloading, which advocates third-party owners of network resource on unlicensed/licensed spectrum to share their spectrum and provide data offloading services, is considered a promising solution to severe spectrum shortage faced by cellular network service providers. In this paper, we consider a general mobile traffic offloading system that adopts the widely used Gale-Shapley algorithm to optimize its mobile phone users (MUs) to offloading stations allocation plan. We notice that without careful protection, such a system could cause serious threat to MUs' location privacy, and thus design effective countermeasures based on the powerful state-of-the-art differential privacy concept. Specifically, we have proposed two joint differentially private Gale-Shapley mechanisms with strong privacy protections for mobile traffic offloading systems. The first mechanism is able to protect each user's location privacy even when all other users collude against this user assuming the system administrator can be trusted. The second mechanism is able to achieve the same privacy guarantee against colluding users, and moreover against an untrusted semi-honest system administrator. We perform extensive experiments to evaluate our mechanisms, and the results show that our mechanisms have good efficiency, accuracy, and privacy protection.

In the problem of privacy-preserving collaborative data publishing, a central data publisher is responsible for aggregating sensitive data from multiple parties and then anonymizing it before publishing for data mining. In such scenarios, the data users may have a strong demand to measure the utility of the published data, since most anonymization techniques have side effects on data utility. Nevertheless, this task is non-trivial, because the utility measuring usually requires the aggregated raw data, which is not revealed to the data users due to privacy concerns. Furthermore, the data publishers may even cheat in the raw data, since no one, including the individual providers, knows the full data set. In this paper, we first propose a privacy-preserving utility verification mechanism based upon cryptographic technique for DiffPart-a differentially private scheme designed for set-valued data. This proposal can measure the data utility based upon the encrypted frequencies of the aggregated raw data instead of the plain values, which thus prevents privacy breach. Moreover, it is enabled to privately check the correctness of the encrypted frequencies provided by the publisher, which helps detect dishonest publishers. We also extend this mechanism to DiffGen-another differentially private publishing scheme designed for relational data. Our theoretical and experimental evaluations demonstrate the security and efficiency of the proposed mechanism.

Pollution attacks can cause severe damages in network coding systems. Many approaches have been proposed to defend against pollution attacks. However, the current approaches implicitly assume that the defender has adequate resources to defend against pollution attacks. When the resources of the defender are limited, they provide no information for the defender to allocate the resources to get better defense performance. In this paper, we consider the case that the defender's resources are limited and study how the defender allocates resources to defend against pollution attacks. We first study this problem in one-session transmissions, and we propose a two-player strategic game to model the interactions between the defender and the attacker. Under this model, two algorithms are proposed to find the best response strategy for the defender. Then, we study the resource allocation problem in a multi-session setting. We propose an extensive game model and an enhancement algorithm to solve the resource allocation problem under this circumstance. Finally, we conducted extensive simulations to evaluate the proposed algorithms. The results demonstrate that our algorithms can significantly improve the utility of the defender, with reasonable computation time.

Mobile phone sensing provides a promising paradigm for collecting sensing data and has been receiving increasing attention in recent years. Different from most existing works, which protect participants' privacy by hiding the content of their data and allow the aggregator to compute some simple aggregation functions, we propose a new approach to protect participants' privacy by delinking data from its sources. This approach allows the aggregator to get the exact distribution of the data aggregation and, therefore, enables the aggregator to efficiently compute arbitrary/complicated aggregation functions. In particular, we first present an efficient protocol that allows an untrusted data aggregator to periodically collect sensed data from a group of mobile phone users without knowing which data belong to which user. Assume there are n users in the group. Our protocol achieves n-source anonymity in the sense that the aggregator only learns that the source of a piece of data is one of the n users. Then, we consider a practical scenario where users may have different source anonymity requirements and provide a solution based on dividing users into groups. This solution optimizes the efficiency of data aggregation and meets all users' requirements at the same time.

To protect individuals' location privacy, an important privacy protection technique that can be used is k -anonymity, which requires at least k users to participate in an anonymity set, so that any user in the set cannot be distinguished from the other k-1 users. However, a significant part of users may not be concerned about their location privacy and therefore may not be interested in participating in the anonymity set. Hence, a prerequisite for achieving k-anonymity location privacy is to stimulate users to participate. In this paper, we revisit the problem of stimulating users that are privacy-indifferent to participate in the anonymity set and providing k-anonymity location privacy for privacy-sensitive users. We first study the case where all privacy-sensitive users have the same requirement of privacy. Then, we extend our study to a more general setting, where privacy-sensitive users have different requirements. For both cases, we design auction-based mechanisms and rigorously prove that the mechanisms are truthful. More importantly, our mechanisms can achieve higher satisfaction ratio than the existing work, i.e., our mechanisms greatly increase the number of privacy-sensitive users successfully winning the auction and receiving privacy protection. We evaluate our mechanisms by using extensive numerical experiments and simulations on a real-world data set. Evaluation results show that our mechanisms achieve much better performance regarding the satisfaction ratio compared with the state-of-the-art mechanisms, and that the computational efficiency is good.

In many existing incentive-based mobile sensing applications, the sensing job owner runs an auction with the mobile phone users to maximize its purchased sensing resource. We notice that both the mobile phone users and the job owner could behave dishonestly to pursue their own interests. This motivates us to design secure and dependable auction mechanisms that generate the correct, promising output even when both of them could cheat. In particular, in this paper, we consider a general auction in which a buyer, who acts as the auctioneer, purchases the resource under a limited budget from a group of sellers who act as the bidders. Considering bidders' privacy and their limited computing capacity, we construct our mechanisms by integrating the innovative game theoretical techniques, logic deductions, and efficient cryptographic operations. Our mechanisms are not only proved to be strategy-proof against dishonest bidders in the sense that they are incentivized to bid their private types truthfully, but also enable all the bidders to efficiently verify the correctness of the auction's outcome, that is computed by the auctioneer, without revealing their private types to each other. Meanwhile, our mechanisms are proved to have the theoretical guarantee that the auctioneer/buyer's expected revenue (i.e. the amount of service it acquires after the auction) is no less than a certain portion of the optimal revenue that the auctioneer can acquire when it knows all the bidders' types at no cost. Our extensive evaluations show that our mechanisms achieve good performance in terms of the revenue maximization and their efficiency.

Pollution attacks can cause severe damages in network coding systems. Many approaches have been proposed to defend against pollution attacks. However, the current approaches implicitly assume that the defender has adequate resources to defend against pollution attacks. When the resources of the defender are limited, they provide no information for the defender to allocate the resources to get better defense performance. In this paper, we consider the case that the defender's resources are limited and study how the defender allocates resources to defend against pollution attacks. We first propose a two-player strategic game to model the interactions between the defender and the attacker. Then, two algorithms are proposed to find the best response strategy for the defender. Finally, we conducted extensive simulations to evaluate the proposed algorithms. The results demonstrate that our algorithms can significantly improve the utility of the defender, with reasonable computation time.

Online behavioral advertising (OBA) has become one of the most successful advertising models on the Internet. Nevertheless, all existing OBA systems are broker-centric in the billing phase, which means it is the broker who exclusively determines advertisers' expenses and publishers' revenues. Consequently, a malicious broker may cheat in their tallying of ad clicks to overcharge advertisers or underpay publishers. Furthermore, as the broker cannot justify the bills, malicious advertisers may deny actual clicks to ask for refunds, and malicious publishers may claim non-existing clicks to demand extra revenue shares. This paper solves these problems by reversing the priority between the broker and the advertisers and publishers. Specifically, when users click on ads, it makes corresponding advertisers and publishers forward click reports of clients to the broker after checking, anonymizing and signing them. The broker then settles accounts with advertisers and publishers fully based on these reports. To guarantee the interests of the broker after the priority reversal, we further propose effective mechanisms for detecting underreporting advertisers and over reporting publishers, respectively.

Today's large-scale enterprise networks, data center networks, and wide area networks can be decomposed into multiple administrative or geographical domains. Domains may be owned by different administrative units or organizations. Hence protecting domain information is an important concern. Existing general-purpose Secure Multi-Party Computation (SMPC) methods that preserves privacy for domains are extremely slow for cross-domain routing problems. In this paper we present PYCRO, a cryptographic protocol specifically designed for privacy-preserving cross-domain routing optimization in Software Defined Networking (SDN) environments. PYCRO provides two fundamental routing functions, policy-compliant shortest path computing and bandwidth allocation, while ensuring strong protection for the private information of domains. We rigorously prove the privacy guarantee of our protocol. We have implemented a prototype system that runs PYCRO on servers in a campus network. Experimental results using real ISP network topologies show that PYCRO is very efficient in computation and communication costs.

To put mobile sensing into large-scale deployments, we have to take care of sensing participants' incentives and privacy first. In this paper, we study how to protect the sensing participants' privacy in the mobile sensing market where multiple sensing jobs reside in one consolidated place. Our problem is highly challenging due to the facts that incentives are introduced and we consider both the sensing job owner and the market administrator could invade the sensing participants' privacy. We propose two privacy-preserving market mechanisms that are able to protect the sensing participants' privacy to solve our problem. Experiments also demonstrate that our mechanisms have good efficiency.

Trajectory data, i.e., human mobility traces, is extremely valuable for a wide range of mobile applications. However, publishing raw trajectories without special sanitization poses serious threats to individual privacy. Recently, researchers begin to leverage differential privacy to solve this challenge. Nevertheless, existing mechanisms make an implicit assumption that the trajectories contain a lot of identical prefixes or n-grams, which is not true in many applications. This paper aims to remove this assumption and propose a differentially private publishing mechanism for more general time-series trajectories. One natural solution is to generalize the trajectories, i.e., merge the locations at the same time. However, trivial merging schemes may breach differential privacy. We, thus, propose the first differentially-private generalization algorithm for trajectories, which leverage a carefully-designed exponential mechanism to probabilistically merge nodes based on trajectory distances. Afterwards, we propose another efficient algorithm to release trajectories after generalization in a differential private manner. Our experiments with real-life trajectory data show that the proposed mechanism maintains high data utility and is scalable to large trajectory datasets.

Decoupling of control and data plane in Software Define Networks (SDN) creates significant flexibility in network management. As networks are evolving into a complex multi-domain multi-layer architecture, traffic engineering across multiple domains and layers entails challenges for the control plane, especially when each separate administrative domain does not disclose their network topology and resource information. In this paper, we present a hierarchical controller design over multidomain and multi-layer networks, by adopting a root controller at the top layer. We allow to aggregate network topology and QoS information into a hierarchical Network Information Base (NIB) for the confidentiality concern. Then, we devise a communication protocol, which enables controllers at different layers and domains to work collaboratively on bandwidth allocation by reading to the hierarchical NIB. We also present an improved traffic engineering algorithm by considering bandwidth and delay simultaneously, to maximize the network utilization while respecting max-min fairness. Experiments on a 717-switches 5-domain network topology demonstrate that our proposal could drive the link utilization ratio to more than 85%.

In wireless ad-hoc networks using XOR network coding, intermediate nodes are required to XOR packets whenever possible. Although there are lots of existing incentive compatible schemes aiming to provide incentives for intermediate nodes to forward packets, they are not suitable for XOR network coding. In this paper, we first present a basic payment scheme to provide incentives to intermediate nodes to follow XOR network coding protocol. We prove that, under our scheme each intermediate node has incentives to follow the XOR protocol. Then we consider the overpayment in the basic scheme, and propose an enhanced payment scheme using SVM model to reduce the payment to a reasonable level. The results show that under the enhanced scheme, the payments can be reduced by up to 70%, and these payment are close to the real forwarding costs.

Collaborative sensing has become increasingly popular in cognitive radio networks to enable unlicensed secondary users to coexist with the licensed primary users and share spectrum without interference. Despite its promise in performance enhancement, collaborative sensing is still facing a lot of security challenges. The problem of revealing secondary users' location information through sensing reports has been reported recently. Unlike any existing work, in this paper we not only address the location privacy issues in the collaborative sensing process against semi-honest adversaries, but also take the malicious adversaries into consideration. We propose efficient schemes to protect secondary users' report from being revealed in the report aggregation process at the fusion center. We rigorously prove that our privacy-preserving collaborative sensing schemes are secure against the fusion center and the secondary users in semi-honest model. We also evaluate our scheme extensively and verify its efficiency.

This paper presents a resource allocation framework for device-to-device (D2D) communications underlaying uplink MIMO cellular networks. At first, our aim is to address the sum-rate maximization problem of the cellular network with both D2D and cellular users. An algorithm based on pure random search is presented for obtaining the optimal resource allocation without using an exhaustive search. Then, we propose a noncooperative resource allocation game for the joint self-optimization of channel allocation, power control, and precoding of the D2D users in a more practical setting. The feasibility and existence of the pure strategy Nash equilibrium are then established. An iterative algorithm based on best response dynamic is then proposed to determine the feasible pure strategy Nash equilibrium under specific conditions. As the algorithm may not always converge, we devise a strategy refinement mechanism to tackle this issue based on the sum-rate criterion. Simulation results verify our theoretical analysis and findings.

Network coding has been shown to be an effective approach to improve the wireless system performance. However, many security issues impede its wide deployment in practice. Besides the well-studied pollution attacks, there is another severe threat, that of wormhole attacks, which undermines the performance gain of network coding. Since the underlying characteristics of network coding systems are distinctly different from traditional wireless networks, the impact of wormhole attacks and countermeasures are generally unknown. In this paper, we quantify wormholes' devastating harmful impact on network coding system performance through experiments. We first propose a centralized algorithm to detect wormholes and show its correctness rigorously. For the distributed wireless network, we propose DAWN, a Distributed detection Algorithm against Wormhole in wireless Network coding systems, by exploring the change of the flow directions of the innovative packets caused by wormholes. We rigorously prove that DAWN guarantees a good lower bound of successful detection rate. We perform analysis on the resistance of DAWN against collusion attacks. We find that the robustness depends on the node density in the network, and prove a necessary condition to achieve collusion-resistance. DAWN does not rely on any location information, global synchronization assumptions or special hardware/middleware. It is only based on the local information that can be obtained from regular network coding protocols, and thus the overhead of our algorithms is tolerable. Extensive experimental results have verified the effectiveness and the efficiency of DAWN.

Informative gene selection is an important topic in the field of bioinformatics which has attracted intensive interest in recent years. It aims to identify the genes which are differentially expressed in different groups, and thus are informative for the classification between the groups. For this purpose, many micro array experiments have been conducted by various medical institutes on their own sets of patients and test subjects. For those institutes who have conducted experiments regarding the same type of disease, it would be beneficial to all of them if they learn on the union of their data to find the informative genes instead of learn just on their own datasets, since the amount of data each institute holds is very limited. However, in many cases, the institutes are not allowed to share their data with others because micro array datasets contain private information about the patients and test subjects. In this paper, we focus on this problem and propose a privacy preserving algorithm that allows multiple parties to perform the widely used informative gene selection method, the Fisher criterion, on the union of their data, without revealing each party's data to others. Basically, we utilize the homomorphic cryptographic system to protect the data during the calculations. Experimental results on real world datasets show the effectiveness of the proposed method.

Network coding has been shown to be an effective approach to improve the wireless system performance. However, many security issues impede its wide deployment in practice. Besides the well-studied pollution attacks, there is another severe threat, that of wormhole attacks, which undermines the performance gain of network coding. Since the underlying characteristics of network coding systems are distinctly different from traditional wireless networks, the impact of wormhole attacks and countermeasures are generally unknown. In this paper, we quantify wormholes' devastating harmful impact on network coding system performance through experiments. Then we propose DAWN, a Distributed detection Algorithm against Wormhole in wireless Network coding systems, by exploring the change of the flow directions of the innovative packets caused by wormholes. We rigorously prove that DAWN guarantees a good lower bound of successful detection rate. We perform analysis on the resistance of DAWN against collusion attacks. We find that the robustness depends on the node density in the network, and prove a necessary condition to achieve collusion-resistance. DAWN does not rely on any location information, global synchronization assumptions or special hardware/middleware. It is only based on the local information that can be obtained from regular network coding protocols, and thus does not introduce any overhead by extra test messages. Extensive experimental results have verified the effectiveness and the efficiency of DAWN.

We apply several state-of-the-art techniques developed in recent advances of counting algorithms and statistical physics to study the spatial mixing property of the two-dimensional codes arising from local hard (independent set) constraints, including: hard-square, hard-hexagon, read/write isolated memory (RWIM), and non-attacking kings (NAK). For these constraints, the strong spatial mixing would imply the existence of polynomial-time approximation scheme (PTAS) for computing the capacity. The existence of strong spatial mixing and a PTAS were previously known for the hard-square constraint. We show the existence of strong spatial mixing for hard-hexagon and RWIM constraints, and consequently we give PTAS for computing the capacities of these codes. We also show evidence that the strong spatial mixing may not hold for the NAK constraint.

Consider a wireless network in which selfish users compete with each other for usage of spectrum. We view this competition as a spatial spectrum access game. There are two fundamental questions regarding this game: how to converge to the optimal Nash equilibrium, and how to converge fast. To answer these two questions, we apply a technique called Belief Propagation to design algorithms for users in this game, which can guarantee fast convergence to an optimal or nearly optimal pure strategy Nash equilibrium. Specifically, when the interference graph is an undirected tree or a directed acyclic graph, our algorithms can find the optimal Nash equilibrium in linear time. For general undirected interference graph, our algorithm converges fast as long as the game is potential (which is the case for many typical scenarios). For some other typical spatial spectrum access games, our algorithms can provide a good approximation to the optimal Nash equilibrium.

Wireless security has been an active research area since the last decade. A lot of studies of wireless security use cryptographic tools, but traditional cryptographic tools are normally based on computational assumptions, which may turn out to be invalid in the future. Consequently, it is very desirable to build cryptographic tools that do not rely on computational assumptions. In this paper, we focus on a crucial cryptographic tool, namely 1-out-of-2 oblivious transfer. This tool plays a central role in cryptography because we can build a cryptographic protocol for any polynomial-time computable function using this tool. We present a novel 1-out-of-2 oblivious transfer protocol based on wireless channel characteristics, which does not rely on any computational assumption. We also illustrate the potential broad applications of this protocol by giving two applications, one on private communications and the other on privacy preserving password verification. We have fully implemented this protocol on wireless devices and conducted experiments in real environments to evaluate the protocol. Our experimental results demonstrate that it has reasonable efficiency.

In cognitive radio networks, cooperative relay is a new technology that can significantly improve spectrum efficiency. While the existing protocols for cooperative relay are very interesting and useful, there is a crucial problem that has not been investigated: Selfish users may cheat in cooperative relay, in order to benefit themselves. Here by cheating we mean the behavior of reporting misleading channel and payment information to the primary user and other secondary users. Such cheating behavior may harm other users and thus lead to poor system throughput. Given the threat of selfish users' cheating, our objective in this paper is to suppress the cheating behavior of selfish users in cooperative relay. Hence, we design the first cheat-proof scheme for cooperative relay in cognitive radio networks, and rigorously prove that under our scheme, selfish users have no incentive to cheat. Our design and analysis start in the model of strategic game for interactions among secondary users; then they are extended to the entire cooperative relay process, which is modeled as an extensive game. To make our schemes more practical, we also consider two aspects: fairness and system security. Results of extensive simulations demonstrate that our scheme suppresses cheating behavior and thus improves the system throughput in face of selfish users.

Dynamic spectrum auctions have been considered a promising approach to effectively re-distribute spectrum resources in the secondary spectrum market. However, the existing spectrum auctions are limited to allocating spectrum in units of channels. Recently software defined radio technologies make exciting progress in operating radios with variable bandwidths. They push the need for designing more flexible spectrum auction frameworks that allow to allocate spectrum with variable bandwidth to the secondary user. In this paper, we design truthful spectrum auction frameworks in which secondary users can bid for, and then be actually allocated spectra with variable bandwidths. We first present a truthful framework for auctions of variable-bandwidth spectra in single collision domains, which can achieve system efficiency. Then, we propose a similar framework for multiple collision domains and rigorously show that it is also truthful. Results of extensive evaluations demonstrate that both of our spectrum auction frameworks for variable bandwidth are effective.

Dynamic spectrum auctions have been considered a promising approach to effectively re-distribute spectrum resources in the secondary spectrum market. However, the existing spectrum auctions are limited to allocating spectrum in units of channels. Recently software defined radio technologies make exciting progress in operating radios with variable bandwidths. They push the need for designing more flexible spectrum auction frameworks that allow to allocate spectrum with variable bandwidth to the secondary user. In this paper, we design truthful spectrum auction frameworks in which secondary users can bid for, and then be actually allocated spectra with variable bandwidths. Instead of submitting valuations for channels (i.e., numbers) as bids, in our frameworks, each secondary user submits his valuation as a function of the bandwidth of spectrum. We first present a truthful framework for auctions of variable-bandwidth spectra in single collision domains, which can achieve system efficiency. Then, we propose a similar framework for multiple collision domains and rigorously show that it is also truthful. Results of extensive evaluations demonstrate that both of our spectrum auction frameworks for variable bandwidth are effective.

Abstract—Channel assignment is a crucial problem for wireless networks, especially for noncooperative wireless networks, in which nodes are selfish. While there have been a few studies of noncooperative, multiradio channel assignment, most existing studies are restricted to single collision domains only. In this paper, we study the design of incentive-compatible protocols for noncooperative, multiradio channel assignment in multiple collision domains. First, we show the necessity of designing incentive-compatible protocols for this problem. Specifically, we show that, if no incentive-compatible protocol is deployed, Nash Equilibria (NEs) may have undesired properties, such as Pareto suboptimality and low throughput. To prevent the system from converging to the NEs with undesired properties, we propose an incentive-compatible protocol for channel assignment in multiple collision domains. We rigorously show that our protocol guarantees that the system converges to NEs that are Pareto-optimal and have the maximum system-wide throughput. Our simulation results also verify that our protocols are effective in ensuring that the system converges to the desired NEs. Index Terms—Wireless access, channel assignment, mechanism design Ç 1

Opportunistic networking is an important technique to enable users to communicate in an environment where contemporaneous end-to-end paths are unavailable or unstable. To support end-to-end messaging in opportunistic networks, a number of probabilistic routing protocols have been proposed. However, when nodes are selfish, they may not have incentives to participate in probabilistic routing, and the system performance will degrade significantly. In this paper, we present novel incentive schemes for probabilistic routing that stimulates selfish nodes to participate. We not only rigorously prove the properties of our schemes, but also extensively evaluate our schemes using GloMoSim. Evaluation results show that there is an up to 75.8% gain in delivery ratio compared with a probabilistic routing protocol providing no incentive.

Gateways are a crucial part of wireless networks. In multi-domain wireless networks, the existing solution to the problem of optimal gateway selection is based on distributed learning. While such a solution is interesting and useful, it has a fundamental limitation: the learning algorithm may stay for long time in a Nash Equilibrium that does not correspond to the optimal gateway selection. This stay can be so long that people can hardly wait for the convergence of the learning algorithm to the optimal gateway selection. In this paper, we present a systematic study of the gateway selection problem. We distinguish three cases and present an alternative solution to the problem in each of these cases: for public link costs, an algorithm to compute the optimal gateway selection directly; for two domains with private link costs, a cryptographic protocol; for three or more domains with private link costs, a perturbation-based protocol. In all the three cases, our solutions accurately compute the optimal gateway selection within reasonable amounts of time, although we emphasize that our solutions are NOT improvements to the gateway selection solutions based on distributed learning (mainly because our solutions are centralized instead of distributed).

Opportunistic networking is an important technique to enable users to communicate in an environment where contemporaneous end-to-end paths are unavailable or unstable. To support end-to-end messaging in opportunistic networks, a number of probabilistic routing protocols have been proposed. However, when nodes are selfish, they may not have incentives to participate in probabilistic routing, and the system performance will degrade significantly. In this paper, we present a novel incentive scheme for probabilistic routing that stimulates selfish nodes to participate. We not only rigorously prove the properties of our scheme, but also extensively evaluate our scheme using GloMoSim. Evaluation results show that there is an up to 75.8% gain in delivery ratio compared with a probabilistic routing protocol providing no incentive.

Wireless security has been an active research area since the last decade. A lot of studies of wireless security use cryptographic tools, but traditional cryptographic tools are normally based on computational assumptions, which may turn out to be invalid in the future. Consequently, it is very desirable to build cryptographic tools that do not rely on computational assumptions. In this paper, we focus on a crucial cryptographic tool, namely 1-out-of-2 oblivious transfer. This tool plays a central role in cryptography because we can build a cryptographic protocol for any polynomial-time computable function using this tool. We present a novel 1-out-of-2 oblivious transfer protocol based on wireless channel characteristics, which does not rely on any computational assumption. We also illustrate the potential broad applications of this protocol by giving an application on private communications. We have fully implemented this protocol on wireless devices and conducted experiments in real environments to evaluate the protocol and its application to private communications. Our experimental results demonstrate that it has reasonable efficiency.

In cognitive radio networks, cooperative relay is a new technology that can significantly improve spectrum efficiency and system throughput. While the existing protocols for cooperative relay are very interesting and useful, there is a crucial problem that has not been investigated: In reality, selfish users may cheat in cooperative relay, in order to benefit themselves. Here by cheating we mean the behavior of reporting misleading information to other users. Such cheating behavior may harm other users and thus lead to poor system throughput.

A wireless ad hoc network does not have an infrastructure, and thus, needs the cooperation of nodes in forwarding other nodes' packets. Reputation system is an effective approach to give nodes incentives to cooperate in packet forwarding. However, existing reputation systems either lack rigorous analysis, or have analysis in unrealistic models. In this paper, we propose FITS, the first reputation system that has rigorous analysis and guaranteed incentive compatibility in a practical model. FITS has two schemes: the first scheme is very simple, but needs a Perceived Probability Assumption (PPA); the second scheme uses more sophisticated techniques to remove the need for PPA. We show that both of these two FITS schemes have a subgame perfect Nash equilibrium in which the packet forwarding probability of every node is one. Experimental results verify that FITS provides strong incentives for nodes to cooperate.

Remote data integrity checking is a crucial technology in cloud computing. Recently, many works focus on providing data dynamics and/or public verifiability to this type of protocols. Existing protocols can support both features with the help of a third-party auditor. In a previous work, Sebé et al. propose a remote data integrity checking protocol that supports data dynamics. In this paper, we adapt Sebé et al.'s protocol to support public verifiability. The proposed protocol supports public verifiability without help of a third-party auditor. In addition, the proposed protocol does not leak any private information to third-party verifiers. Through a formal analysis, we show the correctness and security of the protocol. After that, through theoretical analysis and experimental results, we demonstrate that the proposed protocol has a good performance.

Wireless mesh networks have been widely deployed to provide broadband network access, and their performance can be significantly improved by using a new technology called network coding. In a wireless mesh network using network coding, selfish nodes may deviate from the protocol when they are supposed to forward packets. This fundamental problem of packet forwarding incentives is closely related to the incentive compatible routing problem in wireless mesh networks using network coding, and to the incentive compatible packet forwarding problem in conventional wireless networks, but different from both of them. In this paper, we propose INPAC, the first incentive scheme for this fundamental problem, which uses a combination of game theoretic and cryptographic techniques to solve it. We formally prove that, if INPAC is used, then following the protocol faithfully is a subgame perfect equilibrium. To make INPAC more practical, we also provide an extension that achieves two improvements: (a) an online authority is no longer needed; (b) the computation and communication overheads are reduced. We have implemented and evaluated INPAC on the Orbit Lab testbed. Our evaluation results verify the incentive compatibility of INPAC and demonstrate that it is efficient.

Secure distance-based localization in the presence of cheating beacon (or anchor) nodes is an important problem in mobile wireless ad hoc and sensor networks. Despite significant research efforts in this direction, some fundamental questions still remain unaddressed: In the presence of cheating beacon nodes, what are the necessary and sufficient conditions to guarantee a bounded error during a two-dimensional distance-based location estimation? Under these necessary and sufficient conditions, what class of localization algorithms can provide this error bound? In this paper, we attempt to answer these and other related questions by following a careful analytical approach. Specifically, we first show that when the number of cheating beacon nodes is greater than or equal to a given threshold, there do not exist any two-dimensional distance-based localization algorithms that can guarantee a bounded error. Furthermore, when the number of cheating beacons is below this threshold, we identify a class of distance-based localization algorithms that can always guarantee a bounded localization error. Finally, we outline three novel distance-based localization algorithms that belong to this class of bounded error localization algorithms. We verify their accuracy and efficiency by means of extensive simulation experiments using both simple and practical distance estimation error models.

Channel assignment is a very important topic in wireless networks. In this paper, we study FDMA channel assignment in a noncooperative wireless network, where devices are selfish. Existing work on this problem has considered Nash Equilibrium (NE), which is not a very strong solution concept and may not guarantee a good system performance. In contrast, in this work, we introduce a payment formula to ensure the existence of a Strongly Dominant Strategy Equilibrium (SDSE), a different solution concept that gives participants much stronger incentives. We show that, when the system converges to an SDSE, it also achieves global optimality in terms of system throughput. Furthermore, we extend our work to the case in which some radios have a limited tunability. We show that in such a case, nevertheless, it is generally impossible to have a similar SDSE solution; with additional assumptions on the numbers of radios and the types of channels, etc., we can again achieve an SDSE solution that guarantees optimal system throughput. Besides this extension, we also consider other extensions of our strategic game to achieve throughput fairness and to deal with possibly inconsistent information caused by players joining and leaving. Finally, we evaluate our design with simulated experiments. Numerical results verify that the system does converge to the globally optimal channel assignment with the proposed payment formula, and that the system throughput is significantly higher than that achievable with the random-based and NE-based channel assignment schemes.

Multi-path routing has long been studied as an important routing strategy in networks. Many multi-path routing protocols schedule traffic among multiple paths in order to distribute traffic load. However, existing multi-path routing protocols with traffic assignment require that all nodes in the network follow the protocol, which may not always be a valid assumption when the network consists of selfish nodes. In this paper, we propose an optimal, strategy-proof scheme for multi-path traffic assignment (OSMA) in non-cooperative networks. When OSMA is used, behaving honestly is to the best interest of each selfish node regardless of any other nodes¿ behavior. Furthermore, our scheme is guaranteed to compute the lowest cost traffic assignment with the existence of these selfish nodes. Our evaluations verify that our scheme is optimal and strategy-proof, and demonstrate that the scheme has very low communication and computation overhead.

A body sensor network (BSN) is a network of sensors deployed on a person's body for health care monitoring. Since the sensors collect personal medical data, security and privacy are important components in a BSN. In this paper, we developed IBE-Lite, a lightweight identity-based encryption suitable for sensors in a BSN. We present protocols based on IBE-Lite that balance security and privacy with accessibility and perform evaluation using experiments conducted on commercially available sensors.

With the development of distributed computing environment , many learning problems now have to deal with distributed input data. To enhance cooperations in learning, it is important to address the privacy concern of each data holder by extending the privacy preservation notion to original learning algorithms. In this paper, we focus on preserving the privacy in an important learning model, multilayer neural networks. We present a privacy-preserving two-party distributed algorithm of backpropagation which allows a neural network to be trained without requiring either party to reveal her data to the other. We provide complete correctness and security analysis of our algorithms. The effectiveness of our algorithms is verified by experiments on various real world data sets.

Channel assignment is a very important topic in wireless networks. In this paper, we study FDMA channel assignment in a non-cooperative wireless network, where devices are selfish. Existing work on this problem has considered Nash equilibrium (NE), which is not a very strong solution concept and may not guarantee a good system-wide performance. In contrast, in this work we introduce a payment formula to ensure the existence of a strongly dominant strategy equilibrium (SDSE), a much stronger solution concept. We show that, when the system converges to a SDSE, it also achieves global optimality in terms of effective system-wide throughput. Furthermore, we extend our work to the case in which some radios have limited tunability. We show that, in this case, it is generally impossible to have a similar SDSE solution; but, with additional assumptions on the numbers of radios and the types of channels, etc., we can again achieve a SDSE solution that guarantees globally optimal effective system throughput in the entire system. Besides this extension, we also consider another extension of our strategic game, which is a repeated game that provides fairness. Finally, we evaluate our design in experiments. Our evaluations verify that the system does converge to the globally optimal channel assignment with our designed payment formula, and that the effective system- wide throughput is significantly higher than that of anarchy and Nash equilibrium (NE).

Localization in the presence of malicious beacon nodes is an important problem in wireless networks. Although significant progress has been made on this problem, some fundamental theoretical questions still remain unanswered: in the presence of malicious beacon nodes, what are the necessary and sufficient conditions to guarantee a bounded error during 2-dimensional location estimation? Under these necessary and sufficient conditions, what class of localization algorithms can provide that error bound? In this paper, we try to answer these questions. Specifically, we show that, when the number of malicious beacons is greater than or equal to some threshold, there is no localization algorithm that can have a bounded error. Furthermore, when the number of malicious beacons is below that threshold, we identify a class of localization algorithms that can ensure that the localization error is bounded. We also outline two algorithms in this class, one of which is guaranteed to finish in polynomial time (in the number of beacons providing information) in the worst case, while the other is based on a heuristic and is practically efficient. For completeness, we also extend the above results to the 3-dimensional case. Experimental results demonstrate that our solution has very good localization accuracy and computational efficiency.

Presents the introductory welcome message from the conference proceedings.

User-contributed wireless mesh networks are a disruptive technology that may fundamentally change the economics of edge network access and bring the benefits of a computer network infrastructure to local communities at low cost, anywhere in the world. To achieve high throughput despite highly unpredictable and lossy wireless channels, it is essential that such networks take advantage of transmission opportunities wherever they emerge. However, as opportunistic routing departs from the traditional but less effective deterministic, shortest-path based routing, user nodes in such networks may have less incentive to follow protocols and contribute. In this paper, we present the first routing protocols in which it is incentive-compatible for each user node to honestly participate in the routing despite opportunistic transmissions. We not only rigorously prove the properties of our protocols but also thoroughly evaluate a complete implementation of our protocols. Experiments show that there is a 5.8%-58.0% gain in throughput when compared with an opportunistic routing protocol that does not provide incentives and users can act selfishly.

A body sensor network (BSN), is a network of sensors deployed on a person's body, usually for health care monitoring. Since the sensors collect personal medical data, security and privacy are important components in a body sensor network. At the same time, the collected data has to readily available in the event of an emergency. In this paper, we present IBE-Lite, a lightweight identity-based encryption suitable for sensors, and developed protocols based on IBE-Lite for a BSN.

Multipath routing has long been studied as an important routing strategy in networks. Many multipath routing protocols schedule traffic among multiple paths in order to distribute traffic load. However, existing multipath routing protocols with traffic assignment require that all nodes in the network follow the protocol, which may not always be a valid assumption when the network consists of selfish nodes. In this paper, we propose a traffic assignment scheme to deal with the selfish behavior, which is proved to be strategy-proof. Under our scheme, behaving cooperatively is to the best interest of every node. Extensive evaluations are carried out to show that our scheme has good performance.

In wireless ad hoc networks, routing requires cooperation of nodes. Since nodes often belong to different users, it is highly important to provide incentives for them to cooperate. However, most existing studies of the incentive-compatible routing problem focus on individual nodes' incentives, assuming that no subset of them would collude. Clearly, this assumption is not always valid. In this paper, we present a systematic study of collusion resistance in incentive-compatible routing schemes. In particular, we consider two standard solution concepts for collusion resistance in game theory, namely Group Strategyproofness and Strong Nash Equilibrium. We show that achieving Group Strategyproofness is impossible while achieving Strong Nash Equilibrium is possible. More specifically, we design a scheme that is guaranteed to converge to a Strong Nash Equilibrium. In addition, we give a cryptographic method that prevents profit transfer between colluding nodes, as long as they do not fullytrust each other unconditionally. This method makes our scheme widely applicable in practice. Experiments show that our solution is collusion-resistant and has good performance.

Abnormal events, such as security attacks, misconfigurations, or electricity failures, could have severe consequences toward the normal operation of the Border Gateway Protocol (BGP) that is in charge of the delivery of packets between different autonomous domains, a key operation for the Internet to function. Unfortunately, it has been a difficult task for network security researchers and engineers to classify and detect these events. In our previous work, we have shown that with classification (which relies on the labeling with domain knowledge from BGP experts), it is feasible to effectively detect and distinguish some worms and blackouts from normal BGP behaviors. In this paper, we move one important step forward—we show that we can automatically detect and classify between different abnormal BGP events based on a hierarchy discovered by clustering. As a systematic application of data mining, we devise a clustering method based on normalized BGP data that forms a tree-like hierarchy of abnormal BGP event classes. We then obtain a set of classification rules for each class (node) in the hierarchy, thus able to label unknown BGP data to a closest class. Our method works even as the BGP dynamics evolve over time, as shown in our experiments with seven different abnormal events during a four-year period. Our work, in a more general context, shows it is promising to conduct an interdisciplinary research between network security and data mining in solving real-world problems.

Protection of privacy has become an important problem in data mining. In particular, individuals have become increasingly unwilling to share their data, frequently resulting in individuals either refusing to share their data or providing incorrect data. In turn, such problems in data collection can affect the success of data mining, which relies on sufficient amounts of accurate data in order to produce meaningful results. Random perturbation and randomized response techniques can provide some level of privacy in data collection, but they have an associated cost in accuracy. Cryptographic privacy-preserving data mining methods provide good privacy and accuracy properties. However, in order to be efficient, those solutions must be tailored to specific mining tasks, thereby losing generality. In this paper, we propose efficient cryptographic techniques for online data collection in which data from a large number of respondents is collected anonymously, without the help of a trusted third party. That is, our solution allows the miner to collect the original data from each respondent, but in such a way that the miner cannot link a respondent’s data to the respondent. An advantage of such a solution is that, because it does not change the actual data, its success does not depend on the underlying data mining problem. We provide proofs of the correctness and privacy of our solution, as well as experimental data that demonstrates its efficiency. We also extend our solution to tolerate certain kinds of malicious behavior of the participants. 1.

In many applications, wireless ad-hoc networks are formed by devices belonging to independent users. Therefore, a challenging problem is how to provide incentives to stimulate cooperation. In this paper, we study ad-hoc games---the routing and packet forwarding games in wireless ad-hoc networks. Unlike previous work which focuses either on routing or on forwarding, this paper investigates both routing and forwarding. We first uncover an impossibility result---there does not exist a protocol such that following the protocol to always forward others' traffic is a dominant action. Then we define a novel solution concept called cooperation-optimal protocols. We present Corsac, a cooperation-optimal protocol consisting of a routing protocol and a forwarding protocol. The routing protocol of Corsac integrates VCG with a novel cryptographic technique to address the challenge in wireless ad-hoc networks that a link's cost (ie, its type) is determined by two nodes together. Corsac also applies efficient cryptographic techniques to design a forwarding protocol to enforce the routing decision, such that fulfilling the routing decision is the optimal action of each node in the sense that it brings the maximum utility to the node. Additionally, we extend our framework to a practical radio propagation model where a transmission is successful with a probability. We evaluate our protocols using simulations. Our evaluations demonstrate that our protocols provide incentives for nodes to forward packets.

In order to protect individuals' privacy, the technique of k-anonymization has been proposed to de-associate sensitive attributes from the corresponding identifiers. In this paper, we provide privacy-enhancing methods for creating k-anonymous tables in a distributed scenario. Specifically, we consider a setting in which there is a set of customers, each of whom has a row of a table, and a miner, who wants to mine the entire table. Our objective is to design protocols that allow the miner to obtain a k-anonymous table representing the customer data, in such a way that does not reveal any extra information that can be used to link sensitive attributes to corresponding identifiers, and without requiring a central authority who has access to all the original data. We give two different formulations of this problem, with provably private solutions. Our solutions enhance the privacy of k-anonymization in the distributed scenario by maintaining end-to-end privacy from the original customer data to the final k-anonymous results.

Privacy has become an increasingly important issue in data mining. In this paper, we consider a scenario in which a data miner surveys a large number of customers to learn classification rules on their data, while the sensitive attributes of these customers need to be protected. Solutions have been proposed to address this problem using randomization techniques. Such solutions exhibit a tradeoff of accuracy and privacy: the more each customer's private information is protected, the less accurate result the miner obtains; conversely, the more accurate the result, the less privacy for the customers. In this paper, we propose a simple cryptographic approach that is efficient even in a many-customer setting, provides strong privacy for each customer, and does not lose any accuracy as the cost of privacy. Our key technical contribution is a privacy-preserving method that allows a data miner to compute frequencies of values or tuples of values in the customers’ data, without revealing the privacy-sensitive part of the data. Unlike general-purpose cryptographic protocols, this method requires no interaction between customers, and each customer only needs to send a single flow of communication to the data miner. However, we are still able to ensure that nothing about the sensitive data beyond the desired frequencies is revealed to the data miner. To illustrate the power of our approach, we use our frequency mining computation to obtain a privacy-preserving naive Bayes classifier learning algorithm. Initial experimental results demonstrate the practical efficiency of our solution. We also suggest some other applications of privacy-preserving frequency mining.

Mobile ad hoc networking has been an active research area for several years. How to stimulate cooperation among selfish mobile nodes, however, is not well addressed yet. In this paper, we propose Sprite, a simple, cheat-proof, credit-based system for stimulating cooperation among selfish nodes in mobile ad hoc networks. Our system provides incentive for mobile nodes to cooperate and report actions honestly. Compared with previous approaches, our system does not require any tamper-proof hardware at any node. Furthermore, we present a formal model of our system and prove its properties. Evaluations of a prototype implementation show that the overhead of our system is small. Simulations and analysis show that mobile nodes can cooperate and forward each other's messages, unless the resource of each node is extremely low.