Federated Learning (FL) enables multiple users to collaboratively train a machine learning model without sharing raw data, making it suitable for privacy-sensitive applications. However, local model or weight updates can still leak sensitive information. Secure aggregation protocols mitigate this risk by ensuring that only the aggregated updates are revealed. Among these, single-setup secure aggregation protocols, where key generation and exchange occur only once, are the most efficient due to reduced communication and computation overhead. However, existing single-setup secure aggregation protocols often lack support for dynamic user participation and do not provide strong privacy guarantees such as forward and backward secrecy.

FPGA cloud acceleration, or “FPGA as a Service” (FaaS), offered by AWS, Microsoft Azure, Alibaba Cloud, and Huawei Cloud, has become a promising solution for tackling complex, compute-intensive workloads. It targets applications such as genomics, image and video processing, electronic design automation, compression, and big data analytics. While multi-tenant FPGAs significantly enhances resource utilization efficiency, it faces security threats from power side channels, where attackers craft a malicious circuit to detect voltage fluctuations from victim circuits. Observing that all the crafted circuits exploit either Carry Chain or Look-up Table to sense voltage fluctuations, existing defenses have focused on detecting the malicious use of the two basic FPGA computing resources. However, it remains unclear whether such countermeasures are sufficient to address the growing threat of power side channels in multi-tenant FPGAs. In this paper, we reveal MUXLeak, a novel on-chip sensor that exploitsMultiplexer (MUX)to craft a stealthy power side channel, which bypasses existing countermeasures. Particularly, we perform a thorough analysis of basic resources within an FPGA unit and unveil thatMUX, another basic resource,has never been exploited before. More importantly, it can be directly initialized on Xilinx FPGAs and its incurred signal propagation delay demonstrates an inverse correlation with changes in voltage, making itself exploitable for a new power side channel leakage. In our evaluation, we test MUXLeak on three Xilinx FPGA products and use TDC [18] (i.e., the most sensitive on-chip sensor until now) to benchmark the sensitivity of MUXLeak. Our results show that MUXLeak has achieved the same level of sensitivity as TDC to voltage fluctuations. Further, we apply MUXLeak to mount two attacks, i.e., extracting AES keys within 2.54 hours and stealing DNN model architectures with an accuracy of over 90%.

We unveil discernible temporal (or historical) trajectory imprints resulting from adversarial example (AE) attacks. Standing in contrast to existing studies, which focus on spatial (or static) imprints within the targeted underlying victim models, we present a novel temporal paradigm for understanding these attacks. These imprints are encapsulated within a single loss metric, spanning universally across diverse tasks such as classification and regression, and modalities including image, text, and audio. Recognizing the distinct nature of loss between adversarial and clean examples, we exploit this temporal imprint for AE detection by proposing (Traceable Adversarial Temporal Imprints). TRAIT operates under minimal assumptions without prior knowledge of attacks, thereby framing the detection challenge as a one-class classification problem. However, detecting AEs is still challenged by significant overlaps between the constructed synthetic losses of adversarial and clean examples due to the absence of ground truth for incoming inputs. TRAIT addresses this challenge by converting the synthetic loss into a spectrum signature, using the technique of Fast Fourier Transform to highlight the discrepancies, drawing inspiration from the temporal nature of the imprints, analogous to time-series signals. Across 12 AE attacks including SMACK (USENIX Sec'2023), TRAIT demonstrates consistent outstanding performance across comprehensively evaluated modalities (image, text, audio), tasks (classification and regression), datasets (nine datasets), and model architectures (e.g., ResNeXt50, BERT, RoBERTa, AudioNet). In all scenarios, TRAIT achieves an AE detection accuracy exceeding 97%, often around 99%, while maintaining a false rejection rate of 1%. TRAIT remains effective under the formulated strong adaptive attacks.

As cloud storage is extensively utilized in the contemporary digital age, assuring data integrity and conserving cloud storage space has become a priority for all. However, existing cross-user deduplication audit schemes conflict with the pay-as-you-go model, causing unnecessary costs and violating data isolation. Moreover, retaining a single copy of identical data across multiple users introduces maintenance challenges during data operations. To address these issues, we propose a new blockchain-enabled efficient deduplication and mixed auditing scheme which intricately integrates Message-Locked Encryption (MLE) to construct Homomorphic Verifiable Tags (HVTs), enabling deduplication without exposing confidential data. Our scheme supports single-user deduplication at both block and file levels, as well as plaintext-ciphertext mixed auditing, thereby preventing redundant payments while preserving data isolation to simplify maintenance during data operations and ownership transfers. By employing Elliptic Curve Cryptography (ECC) to encrypt keys and storing the encrypted keys on the blockchain, we ensure data confidentiality while reducing the burden of local key management. Leveraging blockchain-based smart contracts, we further design a self-auditing mechanism that eliminates reliance on trusted third-party auditors. Moreover, our scheme embraces dynamic data operations through an optimized Merkle Hash Tree (MHT) and enables secure cloud data ownership transfer via identity verification. Finally, we prove the correctness and security of our scheme and evaluate its performance through experiments and comparisons with state-of-the-art works, demonstrating its efficiency, particularly in the data upload phase.

Transformer-based Text Classifiers (TTC), exemplified by models like BERT, Roberta, T5, and GPT, have showcased impressive proficiency in Natural Language Processing (NLP). Nonetheless, their vulnerability to adversarial examples (AEs) presents a substantial security concern. Existing efforts to fortify model robustness either entail high computational costs or lack transparency and interpretability. This paper introduces a holistic framework calledExplainability-drivenDetection,Identification, andTransformation (EDIT) to address the shortcomings of prevailing inference-time defenses. (i) EDIT employs explainability methodologies (e.g., attention maps, integrated gradients) and frequency features for automated detection and identification of perturbed words while providing insights into the model's decision-making process. (ii) Building upon detection, EDIT purifies detected adversarial examples through an optimal transformation mechanism that utilizes pre-trained embeddings and model feedback to find the best replacements for identified perturbed words. (iii) To integrate human intelligence, the framework automates the process and triggers alerts for a security analyst to ensure secure safeguards and enable proactive human feedback. We comprehensively evaluate our framework on BERT and ROBERTA, trained on four SOTA text classification datasets: IMDB, YELP, AGNEWS, and SST2. We also assess our framework against seven SOTA Word Substitution Attacks (WSA) at character, word, and multi-levels of granularity. The results show that EDIT achieves an average F-score and balanced accuracy of 89.69% and 89.70%, respectively, across all datasets and classifiers. Compared to four SOTA inference defenses, our detector is 1.22 times better in balanced accuracy (BAL_ACC) and 1.33 times better in F1-Score. It also excels in operational efficiency, being about 83 times faster in feature extraction. The identification module outperforms techniques like ReplaceScore, FreqScore, and ExplainScore, achieving 1.125 times better BAL_ACC and Recall, 1.142 times higher AUC, 1.120 times greater Precision, and 1.149 times improved F1-Score. It significantly reduces false positives and negatives by approximately 1.537 and 1.172 times, respectively. Our transformation module effectively converts adversarial examples into non-adversarial counterparts with an accuracy of 91%. EDIT identifies the necessity for human intervention with an average median accuracy of 89%. With alerts integrated, EDIT achieves an average median accuracy of 90% across all evaluated datasets, models, and types of attacks, all while maintaining computational efficiency, averaging 6.94 seconds.

Federated Learning (FL) enables collaborative machine learning training while preserving data privacy. However, reliance on a central server of the typical FL confronts the risk of single server failure. Decentralized Federated Learning (DFL) emerges as a promising distributed framework, allowing clients to directly share models without server intervention, thereby addressing this challenge. Nevertheless, due to its decentralized nature, DFL is highly susceptible to Byzantine attacks orchestrated by malicious clients. Existing Byzantine-resilient DFL algorithms, though few, remain vulnerable to adaptive attacks due to their heavy reliance ongradient checksof local models, which can be adaptively manipulated by intelligent adversaries. To tackle this issue, we propose a DFL aggregation scheme called FORCE (Byzantine-Resilient Decentralized Federated Learning via Game-Theoretic Contribution Aggregation). Drawing inspiration from the Shapley value in game theory, FORCE shifts from gradient-checking approaches to employ a universal metric, the loss of the local model—independent of specific gradients, to identify potentially malicious clients. Moreover, to reduce the computational overhead of FORCE as the number of neighboring clients scales up, we propose a computationlightweight variant, FORCE−, which is optimized through approximating Shapley value computation. This variant becomes more scalable for resource-restricted DFL clients that are also aggregators. Experimental results on three diverse datasets (two image modalities and one textual modality datasets) under three existing attacks demonstrate that FORCE outperforms existing state-of-the-art Byzantine-resilient DFL aggregation methods, effectively defending against Byzantine attacks.

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

The exploration of backdoor vulnerabilities in object detectors, particularly in real-world scenarios, remains limited. A significant challenge lies in the absence of a natural physical backdoor dataset, and constructing such a dataset is both time- and labor-intensive. In this work, we address this gap by creating a large-scale dataset comprising approximately 11,800 images/frames with annotations featuring natural objects (e.g., T-shirts and hats) as triggers to incur cloaking adversarial effects in diverse real-world scenarios. This dataset is tailored for the study of physical backdoors in object detectors. Leveraging this dataset, we conduct a comprehensive evaluation of an insidious cloaking backdoor effect against object detectors, wherein the bounding box around a person vanishes when the individual is near a natural object (e.g., a commonly available T-shirt) in front of the detector. Our evaluations encompass three prevalent attack surfaces: data outsourcing, model outsourcing, and the use of pretrained models. The cloaking effect is successfully implanted in object detectors across all three attack surfaces. We extensively evaluate four popular object detection algorithms (anchor-based Yolo-V3, Yolo-V4, Faster R-CNN, and anchor-free CenterNet) using 19 videos (totaling approximately 11,800 frames) in real-world scenarios. Our results demonstrate that the backdoor attack exhibits remarkable robustness against various factors, including movement, distance, angle, non-rigid deformation, and lighting. In data and model outsourcing scenarios, the attack success rate (ASR) in most videos reaches 100% or near it, while the clean data accuracy of the backdoored model remains indistinguishable from that of the clean model, making it impossible to detect backdoor behavior through a validation set. Notably, two-stage object detectors (e.g., Faster R-CNN) show greater resistance to backdoor attacks under pure data poisoning conditions (i.e., in data outsourcing) compared to one-stage detectors (e.g., the Yolo series). However, this challenge is surmountable when the attacker controls the training process (particularly in model outsourcing), even with the same small poisoning rate budget as in data outsourcing. In transfer learning attack scenarios assessed on CenterNet, the average ASR remains high at 78%. A detailed 5-minute video illustrating our attack is available at https://youtu.be/Q3HOF4OobbY.

Federated learning (FL) has been widely adopted to provide machine learning (ML) privacy, protecting sensitive user data from leakage. However, there are still attacks that could exploit FL to access users' sensitive data, such as model inversion attacks, property inference attacks, and membership inference attacks. Various solutions were proposed to secure FL using various privacy-preserving techniques, such as differential privacy, homomorphic encryption, and multi-party encryption. However, existing solutions often add noise to the model that hinders the accuracy, or introduce large computational overhead that makes them impractical to use. In this paper, we propose a new privacy protection scheme for FL that uses homomorphic encryption (HE), noise, and secret sharing to protect users' sensitive data from up to n-2 adversarial clients and the server colluding. The computational overhead is minimised by transferring expensive computations of HE to the server, requiring only the encryption and homomorphic addition to be carried out by clients. We provide proof sketches to validate the security of our scheme, and experimental results to demonstrate the practicality of our proposed scheme. The results show that our scheme adds only up to 8% overhead without losing any accuracy to base FL models, showing minimal overhead without losing accuracy, regardless of the data used.

FPGAs offer superior energy efficiency and performance in parallel computing but are vulnerable to remote power side-channel attacks. Existing attacks rely on assumptions of coresident crafted circuits and shared power delivery networks, limiting their practicality in real-world scenarios. In this paper, we present AmpereBleed, a novel current-based side-channel attack that exploits widely available INA226 sensors in ARMFPGA SoCs, bypassing the aforementioned two assumptions. AmpereBleed achieves $261 \times$ greater variations to victim activities compared to the popular ring oscillator (RO) circuit, fingerprints DNN models on the Xilinx Deep Learning Processor Unit (DPU) with $\mathbf{9 9. 7 \%}$ accuracy, and distinguishes the Hamming weights of RSA-1024 keys.

In recent years, cloud providers are dedicated to enabling FPGA multi-tenancy to improve resource utilization, but this new sharing model introduces power side-channel threats, where attackers detect voltage fluctuations from colocated circuits. This paper proposes LeakyDSP, a novel onchip sensor that maliciously configures DSP blocks to sense fine-grained voltage fluctuations but is overlooked by existing studies. Our experimental results show that LeakyDSP achieves high sensitivity to voltage fluctuations and strong robustness to different placements. Besides, we apply LeakyDSP to extract full AES keys with $25 \mathrm{k}-78 \mathrm{k}$ traces and build covert channels with a high transmission rate of 247.94 bit/s.

Federated learning (FL) enables collaborative model training without exposing raw data, offering a privacy-aware alternative to centralized learning. However, FL remains vulnerable to various privacy attacks that exploit shared model updates, including membership inference, property inference, and gradient inversion. Source inference attacks further threaten FL by identifying which client contributed a specific training sample, posing severe risks to user and institutional privacy. Existing source inference attacks mainly assume passive adversaries and overlook more realistic scenarios where the server actively manipulates the training process. In this paper, we present an enhanced source inference attack that demonstrates how a malicious server can amplify behavioral differences between clients to more accurately infer data origin. Our approach introduces active training manipulation and data augmentation to expose client-specific patterns. Experimental results across five representative FL algorithms and multiple datasets show that our method significantly outperforms prior passive attacks. These findings reveal a deeper level of privacy vulnerability in FL and call for stronger defense mechanisms under active threat models.

Graph learning has garnered increasing attention in recent years, which aims to train machine learning models over graph data to support various graph analytic tasks. Coming with the popularity of graph learning are critical privacy concerns regarding the information-rich graphs in many application domains (e.g., finance, social networks, and healthcare). There is thus an urgent call for privacy-preserving graph learning. In this paper, we target an emerging decentralized graph scenario, where a graph is fully decentralized among a set of nodes in such a way that each node only has a limited local view about the global graph. We propose PDGL, a new system framework that can effectively support privacy-assured model training over a decentralized graph, with privacy protection for the links among the nodes as well as the nodes’ private feature data and labels. In contrast to PDGL, prior work does not provide protection for the nodes’ links, feature data, and labels simultaneously. Extensive experiments demonstrate that while providing strong privacy protection for decentralized graph data, PDGL can achieve model utility comparable to the baseline setting of centralized graph learning.

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

With frequent software-induced activations on DRAM rows, bit flips can occur on their physically adjacent rows (i.e., RowHammer). Existing studies leverage FPGA platforms to characterize RowHammer, which have identified key factors that contribute to RowHammer bit flips, e.g., data pattern. As the FPGA-based studies have removed the interference of the OS and the memory controller, their findings on the identified contributing factors do not always work as reported in a real-world computing system, resulting in negative effects on system-level RowHammer attacks and defenses. In this paper, we carry out a system-level empirical study on factors from both the software side and the DRAM side that contribute to RowHammer. We conduct the study on 33 DRAM modules including both DDR4 and DDR3, with 292 DRAM chips from various vendors. Our experimental results from the software side show that some prior findings about existing factors are inconsistent with our observations, thus not applicable to a real-world system. Also, we contribute to identifying one new factor that effectively affects RowHammer bit flips. Our DRAM-side results identify three types of new contributing factors and indicate that DRAM modules are more vulnerable if they achieve better performance and lower power consumption. Particularly, Intel XMP, intended for improving DRAM performance, might be abused for RowHammer attacks.

Protocol Reverse Engineering (PRE) serves as the foundation for numerous security analysis techniques, such as vulnerability mining and intrusion detection, etc. The PRE analysis precision can directly affect the accuracy of these downstream techniques. The network-trace-based PRE technique has become the mainstream PRE technique attributed to its ease of implementation. However, without the prerequisite of additional dedicated devices or knowledge of information, the analysis precision of existing network-trace-based PRE methods is often achievable at only byte or half-byte level but not the fine-grained bit-level, which makes it increasingly challenging to meet the precision requirements of those downstream security applications. In this work, we propose a fine-grained PRE scheme, named FineBID, which makes the identification capability in a fine-grained manner for existing network-trace-based PRE methods into bit-level fields. FineBID follows the global characteristics of protocol fields and constructively models the bit-level field identification problem as a multi-objective decision model, which thus effectively overcomes the insufficient representativeness of bit-level fields’ local characteristics. Then, the multi-objective decision model is solved to obtain the Pareto solution set for different field segmentation levels, and the utility value per bit is further computed. The utility value can be used as the immediate indicator to determine whether each bit is a field boundary or not. Meanwhile, we propose an Actual Ground Truth that is more in line with the actual usage of each bit. With extensive experiments on the Internet, wireless, and industrial protocols, we affirm that FineBID can not only significantly reduce the search space for Ground Truth or Actual Ground Truth with a space reduction of 95.3% compared to exhaustive search, but also identify Ground Truth or Actual Ground Truth more accurately than other similar methods.

Split learning (SL) aims to protect user data privacy by distributing deep models between the client-server and keeping private data locally. In SL training with multiple clients, the local model weights are shared among the clients for the local model updates. This paper first reveals data privacy leakage exacerbated by local weight sharing among the clients in SL through model inversion attacks. Then, to reduce the data privacy leakage issue, we propose and analyze privacy-enhanced SL (P-SL) (or SL without local weight sharing). We further propose a parallelized P-SL to expedite the training process by duplicating multiple server-side model instances without compromising accuracy. Finally, we explore P-SL with late participating clients and devise a server-side cache-based training method to address the forgetting phenomenon in SL when late clients join. The experiment results demonstrate that P-SL helps reduce up to$50\%$of client-side data leakage, which essentially achieves a better privacy-accuracy trade-off than the current trend by using differential privacy mechanisms. Moreover, P-SL and its cache-based version achieve comparable accuracy to baseline SL under various data distributions, while incurring lower costs for computation and communication. Additionally, caching-based training in P-SL mitigates the negative effect of forgetting, stabilizes learning, and enables practical and low-complexity training in a dynamic environment with late-arriving clients.

Model watermarking is a widely adopted mechanism for protecting deep learning (DL) model intellectual property (IP). Black-box verifiable watermarking typically involves injecting backdoors that cause the model to produce predetermined outputs for specific inputs. In contrast, white-box verifiable watermarking uses steganographic techniques to embed watermarks into weight parameters or activation values. However, the former poses new security risks, while the latter often lacks robustness against removal techniques. In this paper, we propose a latent model watermarking, constructing upon the model Division and Union operating concept, dubbed as DUO, leveraging the strengths of two watermarking methods above while eliminating each shortcoming. Once the model owner or provider embeds a watermark into the model using watermark data, the watermarked model is divided into two parts: the main model, which corresponds to the primary task and is made publicly available, and a small sub-network privately reserved by the owner. The watermark resides latently within the main model and can only be activated through the private sub-network (the reserved parameters) when they are united. Consequently, DUO does not adversely affect the performance of the main model on its primary task and does not induce any security risks, even in the presence of watermark data. We extensively validate DUO on four benchmark datasets (CIFAR-10, ImageNette, CIFAR-100, and Tiny-ImageNet) using various model architectures, including standardized ResNet and VGG. The results affirm its capability to accurately verify model ownership without compromising model accuracy. It exhibits a 100% detection accuracy on pirated/positive testing models (96 models are tested) with a 0% false positive rate on normal/negative testing models (64 models are tested). Due to its latent nature, DUO is both effective and robust, capable of withstanding a wide range of state-of-the-art watermark laundering including severe model fine-tuning and pruning. We further evaluate and demonstrate that DUO remains robust against adaptive attacks, even when both the watermark data and the reserved parameters are known to the adversary.

In contrast to prevalent Federated Learning (FL) privacy inference techniques such as generative adversarial networks attacks, membership inference attacks, property inference attacks, and model inversion attacks, we devise an innovative privacy threat: the Data Distribution Decompose Attack on FL, termedDecaf. This attack enables an honest-but-curious FL server to meticulously profile the proportion of each class owned by the victim FL user, divulging sensitive information like local market item distribution and business competitiveness. The crux ofDecaflies in the profound observation that the magnitude of local model gradient changes closely mirrors the underlying data distribution, including the proportion of each class.Decafaddresses two crucial challenges: accurately identify the missing/null class(es) given by any victim user as a premise and then quantify the precise relationship between gradient changes and each remaining non-null class. Notably,Decafoperates stealthily, rendering it entirely passive and undetectable to victim users regarding the infringement of their data distribution privacy. Experimental validation on five benchmark datasets (MNIST, FASHION-MNIST, CIFAR-10, FER-2013, and SkinCancer) employing diverse model architectures, including customized convolutional networks, standardized VGG16, and ResNet18, demonstratesDecaf’s efficacy. Results indicate its ability to accurately decompose local user data distribution, regardless of whether it is IID or non-IID distributed. Specifically, the dissimilarity measured using$L_{\infty }$distance between the distribution decomposed byDecafand ground truth is consistently below 5% when no null classes exist. Moreover,Decafachieves 100% accuracy in determining any victim user’s null classes, validated through formal proof.

While interrupts play a critical role in modern OSes, they have been exploited as a wide range of side channel attacks to break system confidentiality, such as keystroke interrupts, graphic interrupts and network interrupts. However, as previous attacks mainly focus on the exploitation of movable interrupts, they are required to determine which core is handling the target interrupts before their attack, which is non-trivial. The exploitability of non-movable interrupts, which cannot be reassigned by privileged softwares at will, remains unclear. In this paper, we conduct an empirical study on exploitable non-movable interrupts and their contribution to interrupt-based side-channel leakages in x86-based systems. We propose a dynamic analysis technique to investigate how various types of non-movable interrupts are influenced by different workloads. We then conduct a model fingerprinting attack as the benchmark to show that 7 types of non-movable interrupts are exploitable. To demonstrate the viability of these non-movable interrupts, we have created two concrete side channels, called ThermalScope and TimerScope. Specifically, ThermalScope exploits the thermal event interrupts that are triggered only when the CPU temperature exceeds a pre-determined threshold, and TimerScope exploits timer interrupts that are activated regularly to enable the process schedule. Both techniques are adaptable to different attack scenarios, functioning regardless of whether the attacker and victim share the same core or reside on separate cores. Last, we successfully apply them to mount realistic case studies, ranging from constructing cross-core covert channels to breaking kernel address space layout randomization. We also demonstrate successful DNN model fingerprinting attacks under browser scenarios when the frequency scaling is disabled and attacker core is isolated from movable interrupts, where previous HertzBleed, ThermalBleed, and movable interrupt-based attacks are ineffective.

The right to be forgotten has incentivized machine unlearning, but a key challenge persists: the lack of reliable methods to verify unlearning conducted by model providers. This gap facilitates dishonest model providers to deceive data contributors. Current approaches often rely on invasive methods like backdoor injection. However, it poses security concerns and is also inapplicable to legacy data—already released data. To tackle this challenge, this work initializes the first non-invasive unlearning verification framework which operates at triple-granularity (class-, volume-, sample-level) to assess the data facticity and volume integrity of machine unlearning. In this paper, we propose a framework, named TruVRF, encompasses three Unlearning-Metrics, each tailored to counter different types of dishonest model providers or servers (Neglecting Server, Lazy Server, Deceiving Server). TruVRF leverages non-invasive model sensitivity to enable multi-granularity verification of unlearning. Specifically, Unlearning-Metric-I checks if the removed class matches the data contributor’s unlearning request, Unlearning-Metric-II measures the amount of unlearned data, and Unlearning-Metric-III validates the correspondence of a specific unlearned sample with the requested deletion. We conducted extensive evaluations of TruVRF efficacy across three datasets, and notably, we also evaluated the effectiveness and computational overhead of TruVRF in real-world applications for the face recognition dataset. Our experimental results demonstrate that TruVRF achieves robust verification performance: Unlearning-Metric-I and -III achieve over 90% verification accuracy on average against dishonest servers, while Unlearning-Metric-II maintains an inference deviation within 4.8% to 8.2%. Additionally, TruVRF demonstrates generalizability across diverse conditions, including varying numbers of unlearned classes and sample volumes. Significantly, TruVRF is applied to two state-of-theart unlearning frameworks: SISA [3] (presented at Oakland’21) and Amnesiac Unlearning [18], representing exact and approximate unlearning methods, respectively, which affirm TruVRF’s practicality. In addition, we conducted extensive evaluations around TruVRF, including ablation experiments, trade-offs in computational overhead, and the robustness of model sensitivity, among others.

Personal digital data is a critical asset, and governments worldwide have enforced laws and regulations to protect data privacy. Data users have been endowed with the "right to be forgotten" (RTBF) of their data. In the course of machine learning (ML), the forgotten right requires a model provider to delete user data and its subsequent impact on ML models upon user requests. Machine unlearning (MU) emerges to address this, which has garnered ever-increasing attention from both industry and academia. Specifically, MU allows model providers to eliminate the influence of unlearned data without retraining the model from scratch, ensuring the model behaves as if it never encountered this data. While the area has developed rapidly, there is a lack of comprehensive surveys to capture the latest advancements. Recognizing this shortage, we conduct an extensive exploration to map the landscape of MU including the (fine-grained) taxonomy of unlearning algorithms under centralized and distributed settings, debate on approximate unlearning, verification and evaluation metrics, and challenges and solutions across various applications. We also focus on the motivations, challenges, and specific methods for deploying unlearning in large language models (LLMs), as well as the potential attacks targeting unlearning processes. The survey concludes by outlining potential directions for future research, hoping to serve as a beacon for interested scholars.

The proliferation of cloud computing has spurred the popularity of storing and querying databases in the cloud. Among others, skyline queries play an important role in the database field due to its usefulness in multi-criteria decision support systems. To accommodate the tailored needs of users, user-defined skyline query has recently emerged, allowing users to define custom preferences in their skyline queries. However, user-defined skyline query services, if deployed in the cloud, may raise critical privacy concerns as the outsourced databases and skyline queries may contain proprietary/privacy-sensitive information. In light of the above, this paper presents ObliuSky, a new solution enabling oblivious user-defined skyline query processing in the cloud. ObliuSky departs from prior work by not only providing confidentiality protection for the content of the outsourced database, the user-defined skyline queries, and the query results, but also hiding the data patterns (e.g., user-defined dominance relations among database points and search access patterns) which may indirectly cause data leakages. We formally analyze the security guarantees and conduct extensive performance evaluations. The results show that while achieving much stronger security guarantees than the state-of-the-art prior work, ObliuSky is superior in database and query encryption efficiency, and scalable in oblivious query processing.

Spatial crowdsourcing (SC) has recently emerged as a new crowdsourcing service paradigm, where workers move physically to designated locations to perform tasks. Most SC systems perform task assignment based on the spatial proximity between task locations and worker locations. Under such a strategy, workers can only perform tasks near them, which may result in low social welfare (i.e., the total profit of the platform and workers). In contrast, the newly emerging strategy of competitive task assignment (CTA) stimulates workers to compete for their preferred tasks, allowing optimization of the overall profit of SC systems. Among others, one novel CTA setting is competitive detour tasking, which allows workers to compete for tasks that need them to make detours from their original travel paths. However, it requires collecting each worker’s bidding profile which may expose private information. In light of this, in this article, we design, implement, and evaluate PrivCO, a new system framework enabling privacy-preserving competitive detour tasking services in SC. PrivCO delicately bridges state-of-the-art competitive detour tasking algorithms with lightweight cryptography, providing strong protections for workers’ bidding profiles. Extensive experiments over real-world datasets demonstrate that while offering strong security guarantees, PrivCO achieves social welfare comparable to the plaintext domain.

Rowhammer has drawn much attention from both academia and industry in the past years as rowhammer exploitation poses severe consequences to system security. Since the first comprehensive study of rowhammer in 2014, a number of rowhammer attacks have been demonstrated against dynamic random access memory (DRAM)-based commodity systems to break software confidentiality, integrity and availability. Accordingly, numerous software defenses have been proposed to mitigate rowhammer attacks on commodity systems of either legacy (e.g., DDR3) or recent DRAM (e.g., DDR4). Besides, multiple hardware defenses (e.g., Target Row Refresh) from the industry have been deployed into recent DRAM to eliminate rowhammer, which we categorize as production defenses.

All current backdoor attacks on deep learning (DL) models fall under the category of a vertical class backdoor (VCB).In VCB attacks, any sample from a class activates the implanted backdoor when the secret trigger is present, regardless of whether it is a sub-type source-class-agnostic backdoor or a source-class-specific backdoor. For example, a trigger of sunglasses could mislead a facial recognition model when either an arbitrary (source-class-agnostic) or a specific (source-class-specific) person wears sunglasses. Existing defense strategiesoverwhelmingly focus on countering VCB attacks, especially those that are source-class-agnostic. This narrow focus neglects the potential threat of other simpler yet general backdoor types, leading to false security implications. It is, therefore, crucial to discover and elucidate unknown backdoor types, particularly those that can be easily implemented, as a mandatory step before developing countermeasures.

While interrupts play a critical role in modern OSes, they have been exploited as a wide range of side channel attacks to break system confidentiality, such as keystroke interrupts, graphic interrupts and network interrupts. In this paper, we propose ThermalScope, a new side channel that exploits thermal event interrupts, which is adaptable for both native and browser scenarios and incorporates two heat amplifying techniques. The thermal event interrupts are activated only when the CPU package temperature reaches a fixed threshold that is determined by manufacturers. Our key observation is that workloads running on CPUs inevitably generates their distinct heat, which can be correlated with the thermal event interrupts. To demonstrate the viability of ThermalScope, we conduct a comprehensive evaluation on multiple Ubuntu OSes with different Intel-based CPUs. First, we show that the activation of thermal event interrupts correlates with the level of CPU temperature. We then apply ThermalScope to mount different side channel attacks, i.e., building covert channels with a transmission rate of 0.1 b/s, fingerprinting DNN model architectures with an accuracy of over 90% and breaking KASLR within 8.2 hours.

Interrupts are critical hardware resources for OS kernels to schedule processes. As they are related to system activities, interrupts can be used to mount various side-channel attacks (i.e., monitoring keystrokes, inferring website visits, detecting GPU activities, and fingerprinting processes). Given that all these attacks rely on system file interfaces or architectural timers to probe interrupts, various countermeasures have been proposed to either remove the unprivileged access to the file interfaces or detect/cripple architectural timers. In this work, we propose SegScope, a new technique that abuses segment protection to provision fine-grained interrupt observations without any timer. As segment protection is widely used on x86, SegScope works across a wide range of Intel-and AMD-based CPUs. Particularly, we observe that while segment protection preserves the confidentiality of high privileged domain, it leaves a footprint via the data segment registers values when an interrupt occurs. With this key observation, SegScope is crafted by capturing the footprints. To show its security implications, we evaluate it in four case studies. First, SegScope has inferred website visits with a respective success rate of 92.4% on Chrome and 87.4% on Tor Browser in default system settings. Second, SegScope successfully extracts the keys from Cloudflare's Interoperable Reusable Cryptographic Library (CIRCL) vl.l. Third, SegScope steals DNN model architectures with an accuracy of over 80%. Last, SegScope effectively reduces the noise of interrupts to improve the performance of other side channels. As an example, SegScope reduces the error rate of Spectral side channel by 56×. Compared with existing timer-based interrupt-probing techniques, SegScope is fine-grained without introducing false-positives. Further, we leverage SegScope to craft a fine-grained timer, as regular timer interrupts as clock edges contain timestamps. Our evaluation shows that it achieves the same level of timing granularity as the high-resolution timer, i.e., rdtsc and rdpru. We then leverage the timer to break KASLR in about 10 seconds and mount a Flush+Reload based Spectre attack.

Text-to-image diffusion models have achieved tremendous success in the field of controllable image generation, while also coming along with issues of privacy leakage and data copyrights. Membership inference arises in these contexts as a potential auditing method for detecting unauthorized data usage. While some efforts have been made on diffusion models, they are not applicable to text-to-image diffusion models due to the high computation overhead and enhanced generalization capabilities. In this paper, we first identify a conditional overfitting phenomenon in text-to-image diffusion models, indicating that these models tend to overfit the conditional distribution of images given the corresponding text rather than the marginal distribution of images only. Based on this observation, we derive an analytical indicator, namely Conditional Likelihood Discrepancy (CLiD), to perform membership inference, which reduces the stochasticity in estimating memorization of individual samples. Experimental results demonstrate that our method significantly outperforms previous methods across various data distributions and dataset scales. Additionally, our method shows superior resistance to overfitting mitigation strategies, such as early stopping and data augmentation.

Deep Neural Network (DNN) models are often deployed in resource-sharing clouds as Machine Learning as a Service (MLaaS) to provide inference services. To steal model architectures that are of valuable intellectual properties, a class of attacks has been proposed via different side-channel leakage, posing a serious security challenge to MLaaS.Also targeting MLaaS, we propose a new end-to-end attack, DeepTheft, to accurately recover complex DNN model architectures on general processors via the RAPL (Running Average Power Limit)-based power side channel. While unprivileged access to the RAPL has been disabled in bare-metal OSes, we observe that the RAPL is still legitimately accessible in a platform as a service, e.g., the latest docker environment of version 20.10.18 used in this work. However, an attacker can acquire only a low sampling rate (1 KHz) of the time-series energy traces from the RAPL interface, rendering existing techniques ineffective in stealing large and deep DNN models. To this end, we design a novel and generic learning-based framework consisting of a set of meta-models, based on which DeepTheft is demonstrated to have high accuracy in recovering a large number (thousands) of models architectures from different model families including the deepest ResNet152. Particularly, DeepTheft has achieved a Levenshtein Distance Accuracy of 99.75% in recovering network structures, and a weighted average F1 score of 99.60% in recovering diverse layer-wise hyperparameters. Besides, our proposed learning framework is general to other time-series side-channel signals. To validate its generalization, another existing side channel is exploited, i.e., CPU frequency. Different from RAPL, CPU frequency is accessible to unprivileged users in bare-metal OSes. By using our generic learning framework trained against CPU frequency traces, DeepTheft has shown similarly high attack performance in stealing model architectures.

The digitisation of agriculture, integral to Agriculture 4.0, has brought significant benefits while simultaneously escalating cybersecurity risks. With the rapid adoption of smart farming technologies and infrastructure, the agricultural sector has become an attractive target for cyberattacks. This paper presents a systematic literature review that assesses the applicability of existing cyber threat intelligence (CTI) techniques within smart farming infrastructures (SFIs). We develop a comprehensive taxonomy of CTI techniques and sources, specifically tailored to the SFI context, addressing the unique cyber threat challenges in this domain. A crucial finding of our review is the identified need for a virtual Chief Information Security Officer (vCISO) in smart agriculture. While the concept of a vCISO is not yet established in the agricultural sector, our study highlights its potential significance. The implementation of a vCISO could play a pivotal role in enhancing cybersecurity measures by offering strategic guidance, developing robust security protocols, and facilitating real-time threat analysis and response strategies. This approach is critical for safeguarding the food supply chain against the evolving landscape of cyber threats. Our research underscores the importance of integrating a vCISO framework into smart farming practices as a vital step towards strengthening cybersecurity. This is essential for protecting the agriculture sector in the era of digital transformation, ensuring the resilience and sustainability of the food supply chain against emerging cyber risks.

Backdoor attacks on deep learning (DL) models emerge as the most worrisome security threats to their secure and safe usage, especially for security-sensitive tasks. Great efforts have been devoted to thwarting backdoor attacks by devising detection or prevention countermeasures. By default, these countermeasures are designed and evaluated on models with full-precision parameters (e.g., floating32). It is unclear whether they are immediately applicable to mitigate backdoor attacks in the quantized model that are being pervasively deployed on mobile devices and Internet of Things (IoT) devices to save resources (i.e. power and memory) and reduce latency and privacy risks. This work, for the first time, initializes the critical examination of the robustness or applicability of existing state-of-the-art (SOTA) DL backdoor defenses for detecting or preventing backdoor attacks on quantized models. Based on extensive evaluations of four representative defenses (Neural Cleanse, ABS, Fine-Pruning and Trojan Signature) with three datasets (CIFAR10, GTSRB, and STL10), we found that only Neural Cleanse's defensive robustness is generally independent of model quantization, while all others exhibit degraded effectiveness or failures against quantized models (in particular, widely used int-8 and 1-bit models), especially when the model is quantized to be 1-bit. The identified main failure reason is that these defenses are based on examining the weight values of the model or the activation values of the neuron to identify or prevent the backdoor, often using the ranking as a step. Quantization with a small bit width leads to less fine-grained discrete values (e.g., 1-bit quantization only possesses two value elements of -1 and +1), rendering ranking effectiveness deteriorate in this case. Note that the quantization not only applies to the weight but also to activation, thus making these defenses less robust or trivially fail. This work highlights the demand for devising backdoor defenses that are generic to different quantization formats on top of the default full-precision model.

Due to their low latency and high privacy preservation, there is currently a burgeoning demand for deploying deep learning (DL) models on ubiquitous edge Internet of Things (IoT) devices. However, DL models are often large in size and require large-scale computation, which prevents them from being placed directly onto IoT devices, where resources are constrained, and 32-bit floating-point (float-32) operations are unavailable. Commercial framework (i.e., a set of toolkits) empowered model quantization is a pragmatic solution that enables DL deployment on mobile devices and embedded systems by effortlessly post-quantizing a large high-precision model (e.g., float-32) into a small low-precision model (e.g., int-8) while retaining the model inference accuracy. However, their usability might be threatened by security vulnerabilities. This work reveals that standard quantization toolkits can be abused to activate a backdoor. We demonstrate that a full-precision backdoored model which does not have any backdoor effect in the presence of a trigger—as the backdoor is dormant—can be activated by (i) TensorFlow-Lite (TFLite) quantization, the onlyproduct-readyquantization framework to date, and (ii) thebeta releasedPyTorch Mobile framework. In our experiments, we employ three popular model architectures (VGG16, ResNet18, and ResNet50), and train each across three popular datasets: MNIST, CIFAR10 and GTSRB. We ascertain that all trained float-32 backdoored models exhibit no backdoor effecteven in the presence of trigger inputs. Particularly, four influential backdoor defenses are evaluated, and they fail to identify a backdoor in the float-32 models. When each of the float-32 models is converted into an int-8 format model through the standard TFLite or PyTorch Mobile framework's post-training quantization, the backdoor is activated in the quantized model, which shows a stable attack success rate close to 100% upon inputs with the trigger, while it usually behaves upon non-trigger inputs. This work highlights that a stealthy security threat occurs when an end-user utilizes the on-device post-training model quantization frameworks, informing security researchers of a cross-platform overhaul of DL models post-quantization even if these models pass security-aware front-end backdoor inspections. Significantly, we have identified Gaussian noise injection into the malicious full-precision model as an easy-to-use preventative defense against the PQ backdoor. The attack source code is released athttps://github.com/quantization-backdoor.

Byzantine-robust federated learning (FL) endeavors to empower service providers in acquiring a precise global model, even in the presence of potentially malicious FL clients. While considerable strides have been taken in the development of robust aggregation algorithms for FL in recent years, their efficacy is confined to addressing particular forms of Byzantine attacks, and they exhibit vulnerabilities when confronted with a spectrum of attack vectors. Notably, a prevailing issue lies in the heavy reliance of these algorithms on the examination of local model gradients. It is worth noting that an attacker possesses the ability to manipulate a carefully chosen small gradient of a model within a context where there could be millions of gradients available, thereby facilitating adaptive attacks. Drawing inspiration from the foundational Shapley value methodology in game theory, we introduce an effective FL scheme namedCareFL. This scheme is designed to provide robustness against a spectrum of state-of-the-art Byzantine attacks. Unlike approaches that rely on the examination of gradients,CareFLemploys a universal metric, the loss of the local model—independent of specific gradients, to identify potentially malicious clients. Specifically, in each aggregation round, the FL server trains a reference model using a small auxiliary dataset— the auxiliary dataset can be removed with a slight defense degradation trade-off. It employs the Shapley value to assess the contribution of each client-submitted model in minimizing the global model loss. Subsequently, the server selects client models closer to the reference model in terms of Shapley values for the global model update. To reduce the computational overhead ofCareFLwhen the number of clients is relatively scaled-up, we construct its variant, namelyCareFL+ generally by grouping clients. Extensive experimentation conducted on well-established MNIST and CIFAR-10 datasets, encompassing diverse model architectures, including AlexNet, demonstrates thatCareFLconsistently achieves accuracy levels comparable to those attained under attack-free conditions when faced with five formidable attacks.CareFLand CareFL+ outperform six existing state-of-the-art Byzantine-robust FL aggregation methods, includingFLTrust, across both IID and non-IID data distribution settings.

To mitigate recent insidious backdoor attacks on deep learning models, advances have been made by the research community. Nonetheless, state-of-the-art defenses are either limited to specific backdoor attacks (i.e., source-agnostic attacks) or non-user-friendly in that machine learning expertise and/or expensive computing resources are required. This work observes that all existing backdoor attacks have an inadvertent and inevitable intrinsic weakness, termed as non-transferability —that is, a trigger input hijacks a backdoored model but is not effective in another model that has not been implanted with the same backdoor. With this key observation, we propose non-transferability enabled backdoor detection to identify trigger inputs for a model-under-test during run-time. Specifically, our detection allows a potentially backdoored model-under-test to predict a label for an input. Moreover, our detection leverages a feature extractor to extract feature vectors for the input and a group of samples randomly picked from its predicted class label, and then compares the similarity between the input and the samples in the feature extractor’s latent space to determine whether the input is a trigger input or a benign one. The feature extractor can be provided by a reputable party or is a free pre-trained model privately reserved from any open platform (e.g., ModelZoo, GitHub, Kaggle) by a user and thus our detection does not require the user to have any machine learning expertise or perform costly computations. Extensive experimental evaluations on four common tasks affirm that our detection scheme has high effectiveness (low false acceptance rate) and usability (low false rejection rate) with low detection latency against different types of backdoor attacks.

Deep learning models with backdoors act maliciously when triggered but seem normal otherwise. This risk, often increased by model outsourcing, challenges their secure use. Although countermeasures exist, their defense against adaptive attacks is under-examined, possibly leading to security misjudgments. This study is the first intricate examination illustrating the difficulty of detecting backdoors in outsourced models, especially when attackers adjust their strategies, even if their capabilities are significantly limited. It is relatively straightforward for attackers to circumvent detection by trivially violating its threat model (e.g., using advanced backdoor types or trigger designs not covered by the detection). However, this research highlights that various leading detection defenses can simultaneously be evaded using simple adaptive strategies, even under their defined threat models and with limited adversary capabilities (e.g., using easily detectable triggers while maintaining a high attack success rate). To be more specific, this study introduces a novel methodology that employs trigger specificity enhancement and training regulation in a symbiotic manner. This approach allows us to evade multiple backdoor detection defenses simultaneously, including Neural Cleanse (Oakland 19’), ABS (CCS 19’), and MNTD (Oakland 21’). These were the detection tools selected for the Evasive Trojans Track of the 2022 NeurIPS Trojan Detection Challenge. Even when applied in conjunction with these defenses under stringent conditions, such as a high attack success rate (> 97%) and the restricted use of the simplest trigger (small white square), our straightforward method garnered the second prize in NeurIPS Trojan Detection Challenge. Notably, for the first time, our adaptive attack successfully evaded other recent state-of-the-art defenses, including FeatureRE (NeurIPS 22’) and Beatrix (NDSS 23’). This study suggests that existing model outsourcing backdoor defenses remain vulnerable to adaptive attacks, and thus, the use of third-party models should be avoided whenever possible.

Since Deep Learning (DL) backdoor attacks have been revealed as one of the most insidious adversarial attacks, a number of countermeasures have been developed with certain assumptions defined in their respective threat models. However, their robustness is currently inadvertently ignored, which can introduce severe consequences, e.g., a countermeasure can be misused and result in a false implication of backdoor detection. For the first time, we critically examine the robustness of existing backdoor countermeasures. As an initial study, we first identify five potential non-robust failure factors including binary classification, poison rate, model complexity, single-model justification, and hyperparameter sensitivity. As exhaustively examining defenses is infeasible, we instead focus on influential backdoor detection-based countermeasures consisting of model-inspection ones including Neural Cleanse (S&P’19), ABS (CCS’19), and MNTD (S&P’21), and data-inspection ones including SCAn (USENIX SECURITY’21) to examine their failure cases under one or more of these factors. Although these investigated countermeasures claim that they work well under their respective threat models, they have inherent unexplored non-robust cases, which are not even rooted from delicate adaptive attacks. We demonstrate how to trivially bypass them aligned with their respective threat models by simply varying the aforementioned factors. Particularly, for each defense, formal proofs or empirical studies are used to reveal its non-robust cases where it is not as robust as it claims or expects. This work highlights the necessity of thoroughly evaluating the robustness of backdoor countermeasures to avoid their misleading security implications in unknown non-robust cases.

Training deep neural networks (DNNs) requires large datasets and powerful computing resources, which has led some owners to restrict redistribution without permission. Watermarking techniques that embed confidential data into DNNs have been used to protect ownership, but these can degrade model performance and are vulnerable to watermark removal attacks. Recently, DeepJudge was introduced as an alternative approach to measuring the similarity between a suspect and a victim model. While DeepJudge shows promise in addressing the shortcomings of watermarking, it primarily addresses situations where the suspect model copies the victim’s architecture. In this study, we introduce DeepTaster, a novel DNN fingerprinting technique, to address scenarios where a victim’s data is unlawfully used to build a suspect model. DeepTaster can effectively identify such DNN model theft attacks, even when the suspect model’s architecture deviates from the victim’s. To accomplish this, DeepTaster generates adversarial images with perturbations, transforms them into the Fourier frequency domain, and uses these transformed images to identify the dataset used in a suspect model. The underlying premise is that adversarial images can capture the unique characteristics of DNNs built with a specific dataset. To demonstrate the effectiveness of DeepTaster, we evaluated the effectiveness of DeepTaster by assessing its detection accuracy on three datasets (CIFAR10, MNIST, and Tiny-ImageNet) across three model architectures (ResNet18, VGG16, and DenseNet161). We conducted experiments under various attack scenarios, including transfer learning, pruning, fine-tuning, and data augmentation. Specifically, in the Multi-Architecture Attack scenario, DeepTaster was able to identify all the stolen cases across all datasets, while DeepJudge failed to detect any of the cases.

As a critical threat to deep neural networks (DNNs), backdoor attacks can be categorized into two types, i.e., source-agnostic backdoor attacks (SABAs) and source-specific backdoor attacks (SSBAs). Compared to traditional SABAs, SSBAs are more advanced in that they have superior stealthier in bypassing mainstream countermeasures that are effective against SABAs. Nonetheless, existing SSBAs suffer from two major limitations. First, they can hardly achieve a good trade-off between ASR (attack success rate) and FPR (false positive rate). Besides, they can be effectively detected by the state-of-the-art (SOTA) countermeasures (e.g., SCAn [40]).

Radio frequency fingerprint identification (RFFI) is a lightweight device authentication technique particularly desirable for power-constrained devices, e.g., the Internet of things (IoT) devices. Similar to biometric fingerprinting, RFFI exploits the intrinsic and unique hardware impairments resulting from manufacturing, such as power amplifier (PA) nonlinearity, to develop methods for device detection and classification. Due to the nature of wireless transmission, received signals are volatile when communication environments change. The resulting radio frequency fingerprints (RFFs) are distorted, leading to low device detection and classification accuracy. We propose a PA nonlinearity quotient and transfer learning classifier to design the environment-robust RFFI method. Firstly, we formalized and demonstrated that the PA nonlinearity quotient is independent of environmental changes. Secondly, we implemented transfer learning on a base classifier generated by data collected in an anechoic chamber, further improving device authentication and reducing disk and memory storage requirements. Extensive experiments, including indoor and outdoor settings, were carried out using LoRa devices. It is corroborated that the proposed PA nonlinearity quotient and transfer learning classifier significantly improved device detection and device classification accuracy. For example, the classification accuracy was improved by 33.3% and 34.5% under indoor and outdoor settings, respectively, compared to conventional deep learning and spectrogram-based classifiers.

Object detection is the foundation of various critical computer-vision tasks such as segmentation, object tracking, and event detection, which can be deployed on pervasive Internet of Things (IoT) and edge devices. A large amount of data is often required to train an object detector with satisfactory accuracy. However, due to the intensive workforce involved with collecting and annotating large datasets, data curation task is often outsourced to a third party (e.g., Amazon Mechanical Turk) or volunteers. This work reveals severe vulnerabilities in this data curation pipeline. We propose TransCAB, the first work to craft clean-annotated images to stealthily implant the backdoor into the object detectors later trained on them by the data curator/user even when the data curator can manually audit the images and fully controls the training process. Existing clean-label poisoned images are only shown in classification tasks but not non-classification tasks, in particular, object detection due to unique challenges faced, generally owing to the complexity of having multiple objects within each frame (image), including the victim and non-victim objects. Furthermore, we demonstrate that the backdoor effect of both cloaking and misclassification are robustly achieved in the wild when the backdoor is activated with inconspicuously natural physical object as trigger (i.e., T-shirt). The efficacy of our TransCAB is ensured by constructively i) applying the image-camouflage attack that abuses the image-scaling function widely used by the deep learning framework (i.e., PyTorch), ii) incorporating the devised clean image replica technique, and iii) combining identified poison data selection criteria given constrained attacking budget. Extensive experi-ments on YOLOv3, YOLOv4, CenterNet, and Faster R-CNN affirm that TransCAB exhibits more than 90% attack success rate under various real-world scenes even when a very small (i.e., 0.14%) dataset fraction is poisoned. In addition, the small set of poisoned images crafted on one detector (i.e., YOLOv3) can be effectively transferred to insert a backdoor on another detector (i.e., CenterNet). A comprehensive video demo is at https://youtu.be/MA7L_LpXkp4, where a poison rate of merely 0.14% is set for YOLOv4 cloaking backdoor and Faster R-CNN misclassification backdoor. Our collected dataset with T-shirt as a natural trigger (about 11,350 frames in total) is open to the public at https://github.com/inconstance/T-shirt-natural-backdoor-dataset, which is the first relatively large-scale natural trigger backdoor dataset.

To improve the modeling resilience of silicon strong physical unclonable functions (PUFs), in particular, the APUFs that yield a very large number of challenge-response pairs (CRPs), a number of composited APUF variants, such as XOR-APUF, interpose-PUF (iPUF), feed-forward APUF (FF-APUF), and OAX-APUF, have been devised. When examining their security in terms of modeling resilience, utilizing multiple information sources, such as power side channel information (SCI) or/and reliability SCI, given a challenge is under-explored, which poses a challenge to their supposed modeling resilience in practice. Building upon multilabel/head deep learning (DL) model architecture, this work proposes multilabel multiside-channel-information-enabled DL attacks (MLMSAs) to thoroughly evaluate the modeling resilience of aforementioned APUF variants. Despite its simplicity, MLMSA can successfully break large-scaled APUF variants, which has not previously been achieved. More precisely, the MLMSA breaks 128-stage 30-XOR-APUF, (9, 9)- and (2, 18)-iPUFs, and$(2,2,30)$-OAX-APUF when CRPs, power SCI, and reliability SCI are concurrently used. It breaks 128-stage 12-XOR-APUF and$(2,2,9)$-OAX-APUF even when only the easy-to-obtain reliability SCI and CRPs are exploited. The 128-stage six-loop FF-APUF and one-loop 20-XOR-FF-APUF can be broken by simultaneously using reliability SCI and CRPs. All these attacks are normally completed within an hour with a standard personal computer. Therefore, MLMSA is a useful technique for evaluating other existing or any emerging strong PUF designs.

Currently, a high demand for on-device deep neural network (DNN) model deployment is limited by the large model size, computing-intensive floating-point operations (FLOPS), and intellectual property (IP) infringements (i.e., easy access to model duplication for the avoidance of license payments). One appealing solution to addressing the first two concerns is model quantization, which reduces the model size and uses integer operations commonly supported by microcontrollers (MCUs usually do not support FLOPS). To this end, a 1-bit quantized DNN model or deep binary neural network (BNN) significantly improves the memory efficiency, where each parameter in a BNN model has only 1 bit. However, BNN cannot directly provide IP protection (in particular, the functionality of the model is locked unless there is a license payment). In this article, we propose a reconfigurable BNN (RBNN) to further amplify the memory efficiency for resource-constrained Internet of Things (IoT) devices while naturally protecting the model IP. Generally, RBNN can be reconfigured on demand to achieve any one of$M$($M>1$) distinct tasks with the same parameter set, thus only a single task determines the memory requirements. In other words, the memory utilization is improved by a factor of$M$. Our extensive experiments corroborate that up to seven commonly used tasks ($M=7$, six of these tasks are image related and the last one is audio) can co-exist (the value of$M$can be larger). These tasks with a varying number of classes have no or negligible accuracy drop-off (i.e., within 1%) on three binarized popular DNN architectures, including VGG, ResNet, and ReActNet. The tasks span across different domains, e.g., computer vision and audio domains validated herein, with the prerequisite that the model architecture can serve those cross-domain tasks. To fulfill the IP protection of an RBNN model, the reconfiguration can be controlled by both a user key and a device-unique root key generated by the intrinsic hardware fingerprint (e.g., SRAM memory power-up pattern). By doing so, an RBNN model can only be used per paid user per authorized device, thus benefiting both the user and the model provider. The source code is released athttps://github.com/LearningMaker/RBNN.

Building hardware security primitives with on-device memory fingerprints is a compelling proposition given the ubiquity of memory in electronic devices, especially for low-end Internet of Things devices for which cryptographic modules are often unavailable. However, the use of fingerprints in security functions is challenged by the small, but unpredictable variations in fingerprint reproductions from the same device due to measurement noise. Our study formulates a novel and pragmatic approach to achieve highly reliable fingerprints from device memories. We investigate the transformation of raw fingerprints into a noise-tolerant space where the generation of fingerprints is intrinsically highly reliable. We derive formal performance bounds to support practitioners to easily adopt our methods for applications. Subsequently, we demonstrate the expressive power of our formalization by using it to investigate the practicability of extracting noise-tolerant fingerprints from commodity devices. Together with extensive simulations, we have employed 119 chips from five different manufacturers for extensive experimental validations. Our results, including an end-to-end implementation demonstration with a low-cost wearable Bluetooth inertial sensor capable of on-demand and runtime key generation, show that key generators with failure rates less than$10^{-6}$can be efficiently obtained with noise-tolerant fingerprints with a single fingerprint snapshot to support ease-of-enrollment.

Swarm attestation, as an important branch of Remote Attestation (RA), enables a trusted party (verifier) to verify the security states of multiple devices (provers) in a large network (swarm) simultaneously via a challenge-response mechanism. However, swarm attestation suffers from significant redundancy overhead since all devices in the swarm need to be attested in each attestation round. Besides, it faces challenges such as verifier-impersonation Denial of Service (DoS) attacks, highly dynamic networks, transient & self-relocating malware, and Time-Of-Check-Time-Of-Use (TOCTOU) attacks. In this paper, considering not only the detection accuracy but also the privacy of swarm owners in real Internet of Things (IoT) scenarios, we propose an Automatic Federated Swarm Attestation scheme (FeSA). Under this scheme, we design a federated-learning-based automatic swarm attestation protocol that enables theverifiersto identify the suspicious devices by a neural network model and then attest them. To the best of our knowledge, this is the first scheme to apply a federated learning method to RA, ruling out the redundancy attestation rounds while preserving data privacy. The FeSA redesigns the interaction model of RA by a challenge-query mechanism to reduce the overhead of an individual device to a constant. In order to evaluate our scheme, we first set up a smart office environment with 12 types of smart IoT devices for real-world data collection up to 21 days. Based on the real dataset, we demonstrate that FeSA can indeed identify the compromised IoT devices while reducing redundancy. We further simulate large-scale swarms of up to 1,000,000 devices to validate the efficiency of FeSA in large-scale swarms. Last, the security analysis proves the ability of FeSA to resist various attacks.

Emerging ultra-low-power tiny scale computing devices run on harvested energy, are intermittently powered, have limited computational capability, and perform sensing and actuation functions under the control of a dedicated firmware operating without the supervisory control of an operating system. Wirelessly updating or patching firmware of such devices is inevitable. We consider the challenging problem of simultaneous and secure firmware updates or patching for a typical class of such devicesComputational Radio Frequency Identification (CRFID) devices. We propose Wisecr, the first secure and simultaneous wireless code dissemination mechanism to multiple devices that prevents malicious code injection attacks and intellectual property (IP) theft, whilst enabling remote attestation of code installation. Importantly, Wisecr is engineered to comply with existing ISO compliant communication protocol standards employed by CRFID devices and systems. We comprehensively evaluate Wisecr's overhead, demonstrate its implementation over standards compliant protocols, analyze its security, implement an end-to-end realization with popular CRFID devices and open-source the complete software package on GitHub.

Rowhammer is a hardware vulnerability in DRAM memory, where repeated access to hammer rows can induce bit flips in neighboringvictim rows. Rowhammer attacks have enabled privilege escalation, sandbox escape, cryptographic key disclosures, etc. A key requirement ofallexisting rowhammer attacks is that an attacker must have access to at least part of an exploitable hammer row. We term such rowhammer attacks as Explicit Hammer. Recently, several proposals leverage the spatial proximity between the accessed hammer rows and the location of the victim rows for a defense against rowhammer. These all aim to deny the attacker's permission to access hammer rows near sensitive data, thus defeating explicit hammer-based attacks. In this paper, we question the core assumption underlying these defenses. We present Implicit Hammer, a confused-deputy attack that causes accesses to hammer rows that the attacker is not allowed to access. It is a paradigm shift in rowhammer attacks since it crosses privilege boundary to stealthily rowhammer an inaccessible row by implicit DRAM accesses. Such accesses are achieved by abusing inherent features of modern hardware and/or software. We propose a generic model to rigorously formalize the necessary conditions to initiate implicit hammer and explicit hammer, respectively. Compared to explicit hammer, implicit hammer can defeat the advanced software-only defenses, stealthy in hiding itself and hard to be mitigated. To demonstrate the practicality of implicit hammer, we have created two implicit hammer's instances, called PThammer and SyscallHammer.

Internet of Things (IoT) devices upload their data into the cloud for storage because of their limited resources. However, cloud storage data has been subject to potential integrity threats, and consequently auditing techniques are demanded to ensure the integrity of stored data. Unfortunately, existing auditing approaches require owners to undertake expensive tag calculations, which is unsuitable for resource-constrained IoT devices. To resolve the issue, we present aFairCloudAuditing proposal by employing theBlockchain (FCAB). We combine certificateless signatures with the designed dynamic structure to constructively offload the cost of tag computation from the IoT device to the introduced fog node, significantly reducing the local burden. Considering that fog nodes may behave dishonestly during auditing, FCAB enables the IoT device to verify the audit result's authenticity by extracting reliable checking records from the blockchain, thereby achieving auditing fairness, which ensures that thehonestcloud and fog node will gain the corresponding reward. Finally, FCAB is proved to satisfy tag unforgeability, proof unforgeability, privacy preserving, and auditing fairness. Experiment evaluations affirm that FCAB is computationally and communicationally efficient and retains a smaller and fixed computation locally at the data processing stage (mainly including tag computation) than existing auditing methods.

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

Training a Deep Learning (DL) model requires proprietary data and computing-intensive resources. To recoup their training costs, a model provider can monetize DL models through Machine Learning as a Service (MLaaS). Generally, the model is deployed at the cloud, while providing a publicly accessible Application Programming Interface (API) for paid queries to obtain benefits. However, model stealing attacks have posed security threats to this model monetizing scheme as they steal the model without paying for future extensive queries. Specifically, an adversary queries a targeted model to obtain input-output pairs and thus infer the model’s internal working mechanism by reverse-engineering a substitute model, which has deprived model owner’s business advantage and leaked the privacy of the model. In this work, we observe that the confidence vector or the top-1 confidence returned from the model under attack (MUA) varies in a relative large degree given different queried inputs. Therefore, rich internal information of the MUA is leaked to the attacker that facilities her reconstruction of a substitute model. We thus propose to leverage adversarial confidence perturbation to hide such varied confidence distribution given different queries, consequentially against model stealing attacks (dubbed as APMSA). In other words, the confidence vectors returned now is similar for queries from a specific category, considerably reducing information leakage of the MUA. To achieve this objective, through automated optimization, we constructively add delicate noise into per input query to make its confidence close to the decision boundary of the MUA. Generally, this process is achieved in a similar means of crafting adversarial examples but with a distinction that the hard label is preserved to be the same as the queried input. This retains the inference utility (i.e., without sacrificing the inference accuracy) for normal users but bounded the leaked confidence information to the attacker in a small constrained area (i.e., close to decision boundary). The later renders greatly deteriorated accuracy of the attacker’s substitute model. As the APMSA serves as a plug-in front-end and requires no change to the MUA, it is thus generic and easy to deploy. The high efficacy of APMSA is validated through experiments on datasets of CIFAR10 and GTSRB. Given a MUA model of ResNet-18 on the CIFAR10, our defense can degrade the accuracy of the stolen model by up to 15% (rendering the stolen model useless to a large extent) with 0% accuracy drop for normal user’s hard-label inference request.

Vertical federated learning (VFL) has recently emerged as an appealing distributed paradigm empowering multi-party collaboration for training high-quality models over vertically partitioned datasets. Gradient boosting has been popularly adopted in VFL, which builds an ensemble of weak learners (typically decision trees) to achieve promising prediction performance. Recently there have been growing interests in using decision table as an intriguing alternative weak learner in gradient boosting, due to its simpler structure, good interpretability, and promising performance. In the literature, there have been works on privacy-preserving VFL for gradient boosted decision trees, but no prior work has been devoted to the emerging case of decision tables. Training and inference on decision tables are different from that in the case of generic decision trees, not to mention gradient boosting with decision tables in VFL. In light of this, we design, implement, and evaluate Privet, the first system framework enabling privacy-preserving VFL service for gradient boosted decision tables. Privet delicately builds on lightweight cryptography and allows an arbitrary number of participants holding vertically partitioned datasets to securely train gradient boosted decision tables. Extensive experiments over several real-world datasets and synthetic datasets demonstrate that Privet achieves promising performance, with utility comparable to plaintext centralized learning.

Federated learning (FL) and split learning (SL) are state-of-the-art distributed machine learning techniques to enable machine learning training without accessing raw data on clients or end devices. However, their comparative training performance under real-world resource-restricted Internet of Things (IoT) device settings remains barely studied. This work provides empirical comparisons of FL and SL in real-world IoT settings regarding (i) learning performance with heterogeneous data distributions and (ii) on-device execution overhead. Our analyses in this work demonstrate that the learning performance of SL is better than FL under an imbalanced data distribution but worse than FL under an extreme non-IID data distribution. Recently, FL and SL are combined to form splitfed learning (SFL) to leverage each of their benefits (e.g., parallel training of FL and lightweight on-device computation requirement of SL). Our work considers FL, SL, and SFL, and mounts them on Raspberry Pi devices to evaluate their performance, including training time, communication overhead, power consumption, and memory usage with resource-restricted IoT devices. Besides evaluations, we apply two optimizations. First, we generalize SFL by carefully examining the possibility of a hybrid type of model training at the server-side. The generalized SFL merges sequential (dependent) and parallel (independent) processes of model training and thus is beneficial to a system with a large scale of IoT devices, specifically at the server-side operations. Second, we propose pragmatic techniques to substantially reduce the communication overhead by up to four times for the SL and (generalized) SFL.

A physical unclonable function (PUF) generates hardware intrinsic volatile secrets by exploiting uncontrollable manufacturing randomness. Although PUFs provide the potential for lightweight and secure authentication for increasing numbers of low-end Internet of Things devices, practical and secure mechanisms remain elusive. We aim to explore simulatable PUFs (SimPUFs) that are physically unclonable but efficiently modeled mathematically through privileged one-time PUF access to address the above problem. Given a challenge, a securely stored SimPUF in possession of a trusted server computes the corresponding response and its bit-specific reliability. Consequently, naturally noisy PUF responses generated by a resource limited prover can be immediately processed by a one-way function (OWF) and transmitted to the server, because the resourceful server can exploit the SimPUF to perform a trial-and-error search over likely error patterns to recover the noisy response to authenticate the prover. Security of trial-and-error reverse (TREVERSE) authentication under the random oracle model is guaranteed by the hardness of inverting the OWF. We formally evaluate the TREVERSE authentication capability with two SimPUFs experimentally derived from popular silicon PUFs.

Trojan attacks on deep neural networks (DNNs) exploit abackdoorembedded in a DNN model that can hijack any input with an attacker’s chosen signature trigger. Emerging defence mechanisms are mainly designed and validated on vision domain tasks (e.g., image classification) on 2D Convolutional Neural Network (CNN) model architectures; a defence mechanism that is general across vision, text, and audio domain tasks is demanded. This work designs and evaluates a run-time Trojan detection method exploitingSTRongIntentionalPerturbation of inputs that is a multi-domain input-agnostic Trojan detection defence acrossVision,Text andAudio domains—thus termed as STRIP-ViTA. Specifically, STRIP-ViTA is demonstratively independent of not only task domain but also model architectures. Most importantly, unlike other detection mechanisms, it requires neither machine learning expertise nor expensive computational resource, which are the reason behind DNN model outsourcing scenario—one main attack surface of Trojan attack. We have extensively evaluated the performance of STRIP-ViTA over: i) CIFAR10 and GTSRB datasets using 2D CNNs for vision tasks; ii) IMDB and consumer complaint datasets using both LSTM and 1D CNNs for text tasks; and iii) speech command dataset using both 1D CNNs and 2D CNNs for audio tasks. Experimental results based on more than 30 tested Trojaned models (including publicly Trojaned model) corroborate that STRIP-ViTA performs well across all nine architectures and five datasets. Overall, STRIP-ViTA can effectively detect trigger inputs with small false acceptance rate (FAR) with an acceptable preset false rejection rate (FRR). In particular, for vision tasks, we can always achieve a 0 percent FRR and FAR given strong attack success rate always preferred by the attacker. By setting FRR to be 3 percent, average FAR of 1.1 and 3.55 percent are achieved for text and audio tasks, respectively. Moreover, we have evaluated STRIP-ViTA against a number of advanced backdoor attacks and compare its effectiveness with other recent state-of-the-arts.

Backdoor attacks cause model misbehaving by first implanting backdoors in deep neural networks (DNNs) during training and then activating the backdoor via samples with triggers during inference. The compromised models could pose serious security risks to artificial intelligence systems, such as misidentifying ‘stop’ traffic sign into ‘80km/h’. In this paper, we investigate the connection characteristic between the backdoor and the trigger in DNNs and observe the fact that the backdoor is implanted via establishing a link between a cluster of neurons, representing the backdoor, and the triggers. Based on this observation, we design LinkBreaker, a new generic scheme for defending against backdoor attacks. In particular, LinkBreaker deploys a neuron consistency check mechanism for identifying compromised neuron set related to the trigger. Then, the LinkBreaker regulates the model to make predictions based on benign neuron set only and thus breaks the link between the backdoor and the trigger. Compared to previous defenses, LinkBreaker offers a more general backdoor countermeasure that is not only effective against input-agnostic backdoors but also source-specific backdoors, which the later can not be defeated by majority of state-of-the-arts. Besides, LinkBreaker is robust against adversarial examples, which, to a large extent, provides a holistic defense against adversarial example attacks on DNNs, while almost all current backdoor defenses do not have such consideration and capability. Extensive experimental evaluations on real datasets demonstrate that LinkBreaker is with high efficacy of suppressing trigger inputs while incurring no noticeable accuracy deterioration on benign inputs.

Due to the strong analytical ability of big data, deep learning has been widely applied to model on the collected data in industrial Internet of Things (IoT). However, for privacy issues, traditional data-gathering centralized learning is not applicable to industrial scenarios sensitive to training sets, such as face recognition and medical systems. Recently, federated learning has received widespread attention, since it trains a model by only sharing gradients without accessing training sets. But existing research works reveal that the shared gradient still retains the sensitive information of the training set. Even worse, a malicious aggregation server may return forged aggregated gradients. In this article, we propose the VFL, a verifiable federated learning with privacy-preserving for big data in industrial IoT. Specifically, we use Lagrange interpolation to elaborately set interpolation points for verifying the correctness of the aggregated gradients. Compared with existing schemes, the verification overhead of VFL remains constant regardless of the number of participants. Moreover, we employ the blinding technology to protect the privacy of the privacy gradients. If no more than$\boldsymbol{n}$-2 of$\boldsymbol{n}$participants collude with the aggregation server, VFL could guarantee the encrypted gradients of other participants not being inverted. Experimental evaluations corroborate the practical performance of the presented VFL with high accuracy and efficiency.

Image-scaling is a typical operation that processes the input image before feeding it into convolutional neural network models. However, it is vulnerable to the newly revealed image-scaling attack. This work presents an image-scaling attack detection framework, Decamouflage, consisting of three independent detection methods: scaling, filtering, and steganalysis, to detect the attack through examining distinct image characteristics. Decamouflage has a pre-determined detection threshold that is generic. More precisely, as we have validated, the threshold determined from one dataset is also applicable to other different datasets. Extensive experiments show that Decamouflage achieves detection accuracy of 99.9% and 98.5% in the white-box and the black-box settings, respectively. We also measured its running time overhead on a PC with an Intel i5 CPU and 8GB RAM. The experimental results show that image-scaling attacks can be detected in milliseconds. Moreover, Decamouflage is highly robust against adaptive image-scaling attacks (e.g., attack image size variances).

The simplicity of deployment and perpetual operation of energy harvesting devices provides a compelling proposition for a new class of edge devices for the Internet of Things. In particular, Computational Radio Frequency Identification (CRFID) devices are an emerging class of battery free, computational, sensing enhanced devices that harvest all of their energy for operation. Despite wireless connectivity and powering, secure wireless firmware updates remains an open challenge for CRFID devices due to: intermittent powering, limited computational capabilities, and the absence of a supervisory operating system. We present,for the first time, asecurewireless code dissemination (SecuCode) mechanism for CRFIDs by entangling adevice intrinsic hardware security primitive—Static Random Access Memory Physical Unclonable Function (SRAM PUF)—to a firmware update protocol. The design of SecuCode: i) overcomes the resource-constrained and intermittently powered nature of the CRFID devices; ii) is fully compatible with existing communication protocols employed by CRFID devices—in particular, ISO-18000-6C protocol; and ii) is built upon a standard and industry compliant firmware compilation and update method realized by extending a recent framework for firmware updates provided by Texas Instruments. We build an end-to-end SecuCode implementation and conduct extensive experiments to demonstrate standards compliance, evaluate performance and security.

Recent years have witnessed the proliferation of the deployment of virtualization techniques. Virtualization is designed to be transparent, that is, unprivileged users should not be able to detect whether a system is virtualized. Such detection can result in serious security threats such as evading virtual machine (VM)-based malware dynamic analysis and exploiting vulnerabilities for cross-VM attacks. The traditional software-based virtualization leaves numerous artifacts/fingerprints, which can be exploited without much effort to detect the virtualization. In contrast, current mainstream hardware-assisted virtualization significantly enhances the virtualization transparency, making itself more transparent and difficult to be detected. Nonetheless, we showcase three new identified low-level inconspicuous features, which can be leveraged by an unprivileged adversary to effectively and stealthily detect the hardware-assisted virtualization. All three features come from the chipset fingerprints, rather than the traces of software-based virtualization implementations (e.g., Xen or KVM). The identified features include i) Translation-Lookaside Buffer (TLB) stores an extra layer of address translations; ii) Last-Level Cache (LLC) caches one more layer of page-table entries; and iii) Level-1 Data (L1D) Cache is unstable. Based on the above features, we develop three corresponding virtualization detection techniques, which are then comprehensively evaluated on three native environments and three popular cloud providers: i) Amazon Elastic Compute Cloud, ii) Google Compute Engine and iii) Microsoft Azure. Experimental results validate that these three adversarial detection techniques are effective (with no false positive) and stealthy (without triggering suspicious system events, e.g., VM-exit) in detecting the above commodity virtualized environments.

Rowhammer is a destructive software-induced DRAM fault, which an attacker can leverage to break system security. Both individual customers and enterprise users (e.g., cloud providers) might refrain from using a computing system if it is vulnerable to rowhammer vulnerability. In this paper, we provide the first end-to-end tool, coined BitMine, that systematically assesses a DRAM chip’s vulnerability to rowhammer bit flips. BitMine is an extension of DRAMDig. As DRAM address mappings are proprietary techniques and critical in inducing rowhammer bit flips, DRAMDig, our prior work, leverages domain knowledge to efficiently and deterministically reverse-engineer DRAM address mappings on Intel machines. By incorporating DRAMDig, BitMine configures three key parameters, i.e., hammer methods, hammer patterns, data patterns, on the effectiveness of finding rowhammer bit flips. BitMine by default implements 13 hammer methods, 4 hammer patterns and 16 data patterns and is extensible to support more. We evaluate DRAMDig and BitMine against multiple machine models that combine different DRAM chips and Intel microarchitectures. Our experiment results show that DRAMDig efficiently uncovers a deterministic DRAM address mapping for each machine model, and every implemented parameter in BitMine has its distinct effectiveness in triggering bit flips for different machine models.

A new collaborative learning, called split learning, was recently introduced, aiming to protect user data privacy without revealing raw input data to a server. It collaboratively runs a deep neural network model where the model is split into two parts, one for the client and the other for the server. Therefore, the server has no direct access to raw data processed at the client. Until now, the split learning is believed to be a promising approach to protect the client's raw data; for example, the client's data was protected in healthcare image applications using 2D convolutional neural network (CNN) models. However, it is still unclear whether the split learning can be applied to other deep learning models, in particular, 1D CNN.

Side-channel analysis (SCA) becomes a serious realistic threat to crypto devices, it is thus imperative to evaluate the resistance of a device to SCA. Side-channel leakage detection aiming to identify the leakage points potentially revealing secrets in side channel signals, is considered as a preliminary step before further security assessment. This work proposes a novel black-box leakage detection approach, which views the side channel as a constant parameter communication channel when it outputs leakage points. The approach distinguishes leakage points by utilizing the kurtosis-based consistency check for channel parameter estimators. To examine the efficiency of this approach, false negative and false positive rates were first quantitatively analyzed by comprehensive experiments. Considering the fact that side-channel leakage can be from multiple channels in practice, we further investigated the applicability of the proposed approach to multi-channel leakage detection. Interestingly, equipped with the proposed detection approach, we correspondingly devised a novel side-channel attack exploiting a kurtosis-based distinguisher. Overall, extensive experiments have validated the efficiencies of our proposed leakage detection method and the novel SCA attack.

Federated learning (FL) and split neural networks (SplitNN) are state-of-art distributed machine learning techniques to enable machine learning without directly accessing raw data on clients or end devices. In theory, such distributed machine learning techniques have great potential in distributed applications, in which data are typically generated and collected at the client-side while the collected data should be processed by the application deployed at the server-side. However, there is still a significant gap in evaluating the performance of those techniques concerning their practicality in the Internet of Things (IoT)-enabled distributed systems constituted by resource-constrained devices. This work is the first attempt to provide empirical comparisons of FL and SplitNN in real-world IoT settings in terms of learning performance and device implementation overhead. We consider a variety of datasets, different model architectures, multiple clients, and various performance metrics. For the learning performance (i.e., model accuracy and convergence time), we empirically evaluate both FL and SplitNN under different types of data distributions such as imbalanced and non-independent and identically distributed (non-IID) data. We show that the learning performance of SplitNN is better than FL under an imbalanced data distribution but worse than FL under an extreme non-IID data distribution. For implementation overhead, we mount both FL and SplitNN on Raspberry Pi devices and comprehensively evaluate their overhead, including training time, communication overhead, power consumption, and memory usage. Our key observations are that under the IoT scenario where the communication traffic is the primary concern, FL appears to perform better over SplitNN because FL has a significantly lower communication overhead compared with SplitNN. However, our experimental results also demonstrate that neither FL or SplitNN can be applied to a heavy model, e.g., with several million parameters, on resource-constrained IoT devices because its training cost would be too expensive for such devices. Source code is released and available: https://github.com/Minki-Kim95/Federated-Learning-and-Split-Learning-with-raspberry-pi.

A recent trojan attack on deep neural network (DNN) models is one insidious variant of data poisoning attacks. Trojan attacks exploit an effective backdoor created in a DNN model by leveraging the difficulty in interpretability of the learned model to misclassify any inputs signed with the attacker's chosen trojan trigger. Since the trojan trigger is a secret guarded and exploited by the attacker, detecting such trojan inputs is a challenge, especially at run-time when models are in active operation. This work builds STRong Intentional Perturbation (STRIP) based run-time trojan attack detection system and focuses on vision system. We intentionally perturb the incoming input, for instance by superimposing various image patterns, and observe the randomness of predicted classes for perturbed inputs from a given deployed model---malicious or benign. A low entropy in predicted classes violates the input-dependence property of a benign model and implies the presence of a malicious input---a characteristic of a trojaned input. The high efficacy of our method is validated through case studies on three popular and contrasting datasets: MNIST, CIFAR10 and GTSRB. We achieve an overall false acceptance rate (FAR) of less than 1%, given a preset false rejection rate (FRR) of 1%, for different types of triggers. Using CIFAR10 and GTSRB, we have empirically achieved result of 0% for both FRR and FAR. We have also evaluated STRIP robustness against a number of trojan attack variants and adaptive attacks.

A physical unclonable function (PUF), like a fingerprint, exploits manufacturing randomness to endow each physical item with a unique identifier. One primary PUF application is the secure derivation of volatile cryptographic keys using a fuzzy extractor (FE) comprising: 1) a secure sketch and 2) an entropy extractor. Although the entropy extractor can be lightweight, the overhead of the secure sketch responsible for correcting naturally noisy PUF responses is usually high. We observe that, in general, response unreliability with respect to an enrolled reference measurement increases with increasing differences between the in-the-field PUF operating condition and the operating condition used in evaluating the enrolled reference response. For the first time, we exploit such an inadvertent but important observation. In contrast to the conventional single reference response enrollment, we propose enrolling multiple reference responses (MRRs) subject to the same challenge but under multiple distinct operating conditions. The critical observation here is that one of the reference operating conditions is likely to be closer to the operating condition of the field deployed PUF, thus resulting in minimizing the expected unreliability when compared to the single reference under the nominal condition. As a consequence, MRR greatly reduces the demand for the expected number of erroneous bits requiring correction and, subsequently, achieves a significant reduction in the error correction overhead. The significant implementation efficiency gains from the proposed MRR method are demonstrated from software implementations of FEs on batteryless resource constraint computational radio frequency identification devices, where realistic PUF data are collected from intrinsic static random access memory PUFs.

Existing strong controlled physical unclonable function (PUF) designs are built to resist modeling attacks and they deal with noisy PUF responses by exploiting error correction logic. These designs are burdened by the costs of the error correction logic and information shown to leak through the associated helper data for assisting error corrections; leaving the design vulnerable to fault attacks or reliability-based attacks. We present a hybrid PUF-finite state machine (PUF-FSM) construction to realize a controlled strong PUF. The PUF-FSM design removes the need for error correction logic and related computation, storage of the helper data and loading it on-chip by only employing error-free responses judiciously determined on demand in the absence of the underlying PUF-an Arbiter PUF-with a large challenge response pair space. The PUF-FSM demonstrates improved security, especially to reliability-based attacks and is able to support a range of applications from authentication to more advanced cryptographic applications built upon shared keys. We experimentally validate the practicability of the PUF-FSM.