Reliable failure detection and causal reasoning are critical in robotic manipulation, as their absence risks robot damage and endangers human safety.Although recent Vision–Language Models (VLMs) are employed to attempt failure detection and causality reasoning, they typically make retrospective assessment only after task completion, and their reasoning accuracy is often limited.To address these issues, we introduce RoboFailRing, which enables timely failure detection during task execution and enhances the reasoning accuracy of VLMs.It achieves rapid failure detection by retrieving a pre-constructed failure memory and returning a similarity-based decision.In addition, by providing grounded failure report to VLMs, it improves the accuracy of their reasoning about the failure causes and repair strategies.We evaluate RoboFailRing on two large-scale simulated datasets comprising over 6,000 failure trajectories and covering 81 distinct manipulation tasks.The results show that the average success rate of out-of-distribution failure detection reaches 80%, while the mean detection time is cut to roughly 50% of the baseline.Moreover, evaluations on real-world systems show an average 35% gain in VLM failure-reasoning accuracy.We make our code publicly available at: https://github.com/DynamicPoet/RoboFailRing.

As the gap between network and CPU speeds rapidly increases, the CPU-centric network stack proves inadequate due to excessive CPU and memory overheads. Though hardware-offloaded network stacks alleviate these issues, they suffer from limited flexibility in both control and data planes. It seems promising to offload network stacks to Smart-NICs to provide high flexibility. However, naive offloading leads to low throughput due to the inherent architectural limitations of widespread off-path SmartNICs. Even simple operations on staged network traffic would overwhelm the limited SmartNIC memory bandwidth. To this end, we design SmartNS, a SmartNIC-centric network stack with software transport programmability and line-rate packet processing capabilities. To tackle the limitations of SmartNIC-induced challenges, we propose a header-only offloading TX path and an unlimited-working-set in-cache processing RX path to minimize memory traffic to fit the wimpy SmartNIC memory bandwidth. To fully utilize the SmartNIC computing resources, we propose a programmable offloading engine to enable cloud providers to offload customized tasks along with the network stack processing. We prototype SmartNS using the widespread Nvidia BlueField-3 SmartNIC, and implement RoCEv2 and Solar transport protocols by leveraging SmartNS's software programmability. SmartNS achieves 2.2× higher throughput than the microkernel-based baseline in block storage disaggregation and 1.3× higher throughput than the hardware-offloaded baseline in KVCache transfer.

Eavesdropping poses a critical threat to the confidentiality and integrity of voice communications. In recent years, techniques have advanced beyond traditional microphone-based methods toward more intelligent approaches, such as leveraging millimeter-wave sensing to detect the subtle vibrations induced by speakers and reconstruct voice information without direct audio capture. Despite their technical feasibility, these methods remain constrained by limited working ranges—typically only several meters—rendering them impractical for real-world stealthy eavesdropping. In this work, we propose MoiréEar, the first long-range passive eavesdropping system based on moiré patterns. The key idea is to exploit the amplification capability of moiré patterns, which amplify the minute vibrations induced by acoustic signals by hundreds of times, enabling long-range eavesdropping. To make the proposed method even more practical and stealthy, we develop new theoretical foundations that relax the strict requirements for generating moiré patterns. Specifically, our approach enables the use of irregular stripe structures (e.g., commonly seen barcodes) instead of standard moiré gratings to generate moiré patterns. We implement our design using a low-cost photodiode instead of cameras, achieving real-time eavesdropping with lightweight signal processing. Comprehensive experiments show that the system can extract intelligible audio at a distance of up to 90 m, outperforming the state of the art by an order of magnitude in range. We believe this new eavesdropping modality can inspire a wide range of IoT applications.

To sustain the exponential growth of AI, large-scale multi-node LLM services are increasingly deployed behind Server Load Balancers (SLBs). While these services heavily rely on Remote Direct Memory Access (RDMA) for high-performance communication, natively integrating RDMA with standard SLB architectures introduces severe incompatibilities, such as centralized node bottlenecks, scheduling inconsistencies, and performance degradation during long-distance RDMA transmission. Hence, we propose Draco-SLB, a novel endpoint-side transport shim layer. Draco-SLB transparently shields underlying RDMA execution from network-side complexities, including SLBs and intermediate network middleboxes.

Recently reported attacks against Programmable Logic Controllers (PLCs) have shown that the exploitation of Industrial Control Protocols (ICPs), i.e., ICP-based attacks, poses significant threats to industrial control systems. ICP-based attacks include two essential steps: generating tailored attack payloads and breaking through the session-ID-based PLC defenses. Traditional approaches to performing the two steps rely on laborious manual analysis. To analyze the threats posed by ICP-based attacks to commercial-off-the-shelf PLCs, we propose AIAF, an Automated ICP-based Attack Framework leveraging proprietary binary protocols, which operates automatically through an offline construction of effective attack payloads and an online ICP-based attack test. We have evaluated AIAF with 9 mainstream PLCs, covering 9 protocols, showing that AIAF can reverse engineer 12 kinds of session-ID negotiation (6 value-changed and 6 value-same), generate attack payloads, and execute 35 ICP-based attacks with a 94.29% success rate. Our further Internet-wide evaluation reveals that over 28K PLCs exposed to the Internet are vulnerable to ICP-based attacks.

Millimeter-wave (mmWave) technology offers significant potential for high-bandwidth, low-latency communication industrial applications. However, mmWave faces several challenges, such as limited range, susceptibility to blockage, and slow beam alignment. For this, mmWave access points (APs) are often deployed at high densities, which leads to increased costs. In this paper, we introduce mmProjector, the first cost-effective reflector, which is non-reconfigurable with a static structure but can serve mobile industrial robots. The key idea is to reshape the reflected waves only along the robots’ movement trajectories using mmProjector and make the phase of reflected wave constructively added along the trajectories, improving reflection gain with limited incident waves. To achieve this, we develop an analytical model grounded in electromagnetic theory and propose sub-optimal algorithms for the deployment and shape design of reflectors. We prototype mmProjector and deploy it in a real-world airplane assembly factory. Experiments demonstrate that mmProjector operates effectively at 60 GHz, delivering up to 1.8 Gbps of bandwidth in non-line-of-sight (NLOS) scenarios, while achieving up to 75% reduction in costs.

Deep reinforcement learning (DRL) has been implemented across various critical applications, including smart grids, trac management systems, and autonomous vehicles. To safeguard intellectual property and mitigate security vulnerabilities, access to DRL models is typically restricted to a black-box format. is means specic details like the structure of the policy network and optimization processes are not openly available to users. It is crucial to determine if the hyper-parameters can be inferred from observable states and actions within these models, presenting two primary challenges: 1) limited data available from the black-box model and 2) the intertwined eects of hyperparameters on the model's behavior. Since DRL models exhibit varying behaviors in identical tasks depending on their hyper-parameter congurations, we introduce a novel hyper-parameter inference attack against DRL, named HyperInfer, which allows adversaries to deduce the settings of a black-box DRL model. In order to fully assess the risk of model hyper-parameter leakage, we design two novel state generation methods that provoke divergent responses from DRL models. We also develop an inference framework to elucidate the relationship between model behavior and hyper-parameter settings. rough comprehensive experiments involving multiple DRL models and environments, we demonstrate that model behaviors can indeed reveal hyper-parameter settings, with inference accuracy surpassing 90% in scenarios such as PPO with CartPole. We also discuss keyndings relevant to practical applications and explore how knowledge of hyperparameters can facilitate more sophisticated attacks. Lastly, we propose potential defensive strategies to minimize the risk of hyper-parameter leakage in DRL models.

Machine unlearning is the primary way to fight for the “right to be forgotten” in machine learning field, which is promoted among multiple privacy legislations, such as GDPR and CCPA. However, the latest work has shown that machine unlearning in deep learning cannot be easily verified, making it challenging for the data owners to be convinced that their data has indeed been deleted as claimed. This is especially problematic for federated learning (FL), where a number of participants jointly train a global model while each participant should be free to join and leave the federation as they wish. However,the lack of a reliable approach to verify unlearning in FL will no doubt discourage certain users from joining the federation.In this work, we propose A2E, a black-box watermarking scheme from a leaving participant's perspective to realize verifiable federated unlearning which incurs minimum impact and no security threats to vanilla FL. The key idea is to leverage adversarial training to inject the anti-adversarial example (A2E) characteristic into the uploaded model updates of the last contribution round as the watermark of the leaving participant. Then, we verify whether the server has indeed executed the effective unlearning, with the newly developed probabilistic quantification of unlearning confidence, by checking the unlearned global model's resistance to the specially generated watermark-dependent adversarial examples of the leaver. We conducted large-scale experiments on various popular datasets (including natural images, medical images, and speech) and model structures (including LeNet, ResNet, VGG, and LSTM). The results confirm the effectiveness of A2E in verifying federated unlearning with a high confidence. We also show that A2E is robust against multiple adaptive strategies from the adversarial server and participants.

Programmable logic controllers (PLCs) play a crucial role in various industrial manufacturing processes. Recent attack events show that attackers have a strong interest in controller variables of PLCs, including the device status and internal program logic. Detecting anomalous messages targeting PLC controller variables, which relies on the analysis of controller variable semantics, has proven to be an effective method for identifying such attacks. However, the proprietary nature of industrial control protocols (ICPs) poses a challenge to extracting the required semantics. In this paper, we propose an automated framework namedSePannerto extract the semantics of controller variables from proprietary ICPs based on network traffic. Specifically, we first collect multiple groups of interaction traffic of PLCs and perform the starting-aligned comparisons on them to locate the semantic fields directly. Then, we identify and investigate a new problem in semantic extraction — interference resulting from misordered messages — and propose a set of filtering criteria to eliminate it effectively. We evaluate SePanner using the S7COMM protocol, and the results indicate that SePanner can successfully extract the semantics of controller variables with 100% accuracy. Additionally, we employ SePanner to analyze 7 proprietary ICPs, successfully extracting the semantics of 63 controller variables and their 134 states. Additionally, we demonstrate the extensive applications of SePanner in multiple ICS security scenarios and present its better performance compared with existing ICP semantic analyzing tools.

Different from traditional object detectors such as YOLO, Detection Transformers (DETR) have reshaped the landscape of object detection by replacing heuristic-driven components like Non-Maximal Suppression with a fully end-to-end framework based on one-to-one Hungarian matching. While the majority of research has focused on improving the slow training convergence of DETR, this work investigates their security from an adversarial perspective. We unveil a critical vulnerability stemming directly from DETR’s core design: the deterministic one-to-one mapping between object queries and ground-truth objects can be exploited. This allows an adversary to craft perturbations that selectively manipulate specific target objects – causing them to vanish or be misclassified – while preserving the detection integrity of all other objects in the scene. Our initial analysis reveals that conventional gradient-based attacks are ill-suited for this task, as they induce unintended interference on non-target instances, a phenomenon we term as the “spillover effect”. To overcome this, we re-formulate the attack optimization by incorporating a novel penalty term that explicitly decouples the adversarial influence on target and non-target objects. Furthermore, we provide theoretical analysis to derive perturbation bounds under which the optimal matching assignments remain invariant, offering deeper insights into the model’s stability. Extensive experiments on standard benchmarks demonstrate that our proposed attack significantly improves the success rate and convergence speed while inducing far fewer feature-level artifacts, making the attack both more effective and stealthier.

In privacy-preserving mobile network transmission scenarios with heterogeneous client data, personalized federated learning methods that decouple feature extractors and classifiers have demonstrated notable advantages in enhancing learning capability. However, many existing approaches primarily focus on feature space consistency and classification personalization during local training, often neglecting the local adaptability of the extractor and the global generalization of the classifier. This oversight results in insufficient coordination and weak coupling between the components, ultimately degrading the overall model performance. To address this challenge, we propose FedeCouple, a federated learning method that balances global generalization and local adaptability at a fine-grained level. Our approach jointly learns global and local feature representations while employing dynamic knowledge distillation to enhance the generalization of personalized classifiers. We further introduce anchors to refine the feature space; their strict locality and non-transmission inherently preserve privacy and reduce communication overhead. Furthermore, we provide a theoretical analysis proving that FedeCouple converges for nonconvex objectives, with iterates approaching a stationary point as the number of communication rounds increases. Extensive experiments conducted on five image-classification datasets demonstrate that FedeCouple consistently outperforms nine baseline methods in effectiveness, stability, scalability, and security. Notably, in experiments evaluating effectiveness, FedeCouple surpasses the best baseline by a significant margin of 4.3%.

Millimeter-wave (mmWave) communication offers a promising solution for high-throughput, low-latency unmanned aerial vehicle (UAV) networks. However, maintaining strong received signal strength (RSS) remains a challenge due to UAV mobility. While existing studies have largely focused on beam alignment, they often overlook another critical issue: polarization mismatch caused by UAV orientation changes. This problem is particularly severe in cost-sensitive commercial off-the-shelf (COTS) mmWave devices, which typically employ linearly polarized (LP) antenna arrays. Our measurements reveal that even with perfect beam alignment, UAV orientation can still cause significant signal degradation due to polarization mismatch. To address this challenge, we propose PolarFix, a practical metasurface solution that enables real-time polarization matching without requiring any modifications to existing transceiver hardware. Specifically, we design a linear-to-circular polarization (L2C) metasurface that transforms linearly polarized (LP) waves into circularly polarized signals, allowing LP antennas to maintain consistent signal power despite changes in UAV orientation.

Most of the federated learning techniques are limited to homogeneous model fusion. With the rapid growth of smart applications on resource-constrained edge devices, it becomes a barrier to accommodate their heterogeneous computing power and memory in the real world. Federated Distillation is a promising alternative to enable aggregation from heterogeneous models. However, the effectiveness of knowledge transfer still remains elusive under the shadow of distinct representation power from heterogeneous models. In this paper, we approach from an adversarial perspective to characterize the decision boundaries during distillation. By leveraging K-step PGD attacks, we successfully model the dynamics of the closest boundary points and establish a quantitative connection between the predictive uncertainty and boundary margin. Based on these findings, we further propose a new loss function to make the distillation attend to samples close to the decision boundaries, thus learning from more informed logit distributions. The extensive experiments over CIFAR-10/100 and Tiny-ImageNet demonstrate about 0.5-3.5% improvement of accuracy under different IID and non-IID settings, with only a small increment of computational overhead.

Object detection is a fundamental enabler for many real-time downstream applications such as autonomous driving, augmented reality and supply chain management. However, the algorithmic backbone of neural networks is brittle to imperceptible perturbations in the system inputs, which were generally known as misclassifying attacks. By targeting the real-time processing capability, a new class of latency attacks has been reported recently. They exploit new attack surfaces in object detectors by creating a computational bottleneck in the post-processing module, which leads to cascading failure and puts the real-time downstream tasks at risk. In this work, we take an initial attempt to defend against this attack via background-attentive adversarial training that is also cognizant of the underlying hardware capabilities. We first draw system-level connections between latency attacks and hardware capacity across heterogeneous GPU devices. Based on the particular adversarial behaviors, we utilize objectness loss as a proxy and build background attention into the adversarial training pipeline, and achieve a favorable balance between clean and robust accuracy. The extensive experiments demonstrate the effectiveness of the defense in restoring real-time processing capability from 13 FPS to 43 FPS on Jetson Orin NX, with a better trade-off between the clean and robust accuracy. The source code is available at: https://github.com/Hill-Wu1998/underload.

Federated semantic segmentation enables pixel-level classification in images through collaborative learning while maintaining data privacy. However, existing research commonly overlooks the fine-grained class relationships within the semantic space when addressing heterogeneous problems, particularly domain shift. This oversight results in ambiguities between class representation. To overcome this challenge, we propose a novel federated segmentation framework that strikes class consistency, termed FedSaaS. Specifically, we introduce class exemplars as a criterion for both local- and global-level class representations. On the server side, the uploaded class exemplars are leveraged to model class prototypes, which supervise global branch of clients, ensuring alignment with global-level representation. On the client side, we incorporate an adversarial mechanism to harmonize contributions of global and local branches, leading to consistent output. Moreover, multilevel contrastive losses are employed on both sides to enforce consistency between two-level representations in the same semantic space. Extensive experiments on five driving scene segmentation datasets demonstrate that our framework outperforms state-of-the-art methods, significantly improving average segmentation accuracy and effectively addressing the class-consistency representation problem.

As the implementation of machine learning (ML) systems becomes more widespread, especially with the introduction of larger ML models, we perceive a spring demand for massive data. However, it inevitably causes infringement and misuse problems with the data, such as using unauthorized online artworks or face images to train ML models. To address this problem, many efforts have been made to audit the copyright of the model training dataset. However, existing solutions vary in auditing assumptions and capabilities, making it difficult to compare their strengths and weaknesses. In addition, robustness evaluations usually consider only part of the ML pipeline and hardly reflect the performance of algorithms in real-world ML applications. Thus, it is essential to take a practical deployment perspective on the current dataset copyright auditing tools, examining their effectiveness and limitations. Concretely, we categorize dataset copyright auditing research into two prominent strands: intrusive methods and non-intrusive methods, depending on whether they require modifications to the original dataset. Then, we break down the intrusive methods into different watermark injection options and examine the non-intrusive methods using various finger-prints. To summarize our results, we offer detailed reference tables, highlight key points, and pinpoint unresolved issues in the current literature. By combining the pipeline in ML systems and analyzing previous studies, we highlight several future directions to make auditing tools more suitable for real-world copyright protection requirements.

Text-to-image models based on diffusion processes, such as DALL-E, Stable Diffusion, and Midjourney, are capable of transforming texts into detailed images and have widespread applications in art and design. As such, amateur users can easily imitate professional-level paintings by collecting an artist's work and fine-tuning the model, leading to concerns about artworks' copyright infringement. To tackle these issues, previous studies either add visually imperceptible perturbation to the artwork to change its underlying styles (perturbation-based methods) or embed post-training detectable watermarks in the artwork (watermark-based methods). However, when the artwork or the model has been published online, i.e., modification to the original artwork or model retraining is not feasible, these strategies might not be viable.

Millimeter-wave (mmWave) communication has great potential for high rates and low latency but suffers from severe non-line-of-sight (NLOS) blockage and high cost of beam alignment. These problems exacerbate for mobile industrial robots. Most existing methods use reconfigurable intelligent surface (RIS) to change the channel environment, whose practical applications are hindered by high cost. This paper introduces low-cost mmFlower, consisting of an array of 3D-printing reflector units whose orientation can be adjusted mechanically. We first construct a theoretical reflection model, bridging the mechanical parameters and communication metrics. mmFlower works in two modes according to whether it needs real-time reconfiguration. On the one hand, based on our uniform-RSS (received signal strength) projection algorithm, mmFlower remains static and reshapes mmWave into arbitrary projection trajectory along the robot movement. It maximizes mmWave allocation to the more frequently visited areas and realizes seamless mmWave coverage. On the other hand, mmFlower can dynamically track mmWave pencil beams on robots, providing higher reflection gain. Extensive evaluations in an airplane assembly factory show the superiority of mmFlower on reflection gain (supporting up to 1.8 Gbps in NLOS), flexibility to different trajectories of robots, robustness to deployment deviation, etc.

Stealthy attacks manipulate the operation of Industrial Control Systems (ICSs) without being undetected, allowing persistent manipulation of system operation and thus the potential to cause destructive damage. This paper introduces a new vulnerability of ICS that can be exploited to mount stealthy attacks without requiring any domain knowledge. This vulnerability is caused by a common practice in system monitoring, i.e., the SCADA monitors ICS operation at a much lower frequency than system execution, causing a loss of precision when the SCADA tries to cross-validate the issued control commands using the collected sensory data. Exploiting this vulnerability, an attack calledPLC-SAGEis designed to stealthily manipulate the system operation by identifying and injecting malicious control commands that will not be concluded as abnormal by the SCADA. This paper further discusses a preferred ICS engineering practice and an attestation strategy to mitigate the above vulnerability and protect ICS fromPLC-SAGE. BothPLC-SAGEand the proposed mitigations have been experimentally validated on two ICS platforms.

The security risk of semantic attacks to Industrial Control Systems (ICSs) is increasing. Semantic attacks manipulate targeted system modules by identifying the physical semantics of variables in Programmable Logic Controllers (PLCs) programs, i.e., the sensing/actuating modules represented by the variables, which is usually and inefficiently achieved via manual examination of system documents and long-term observation of system behavior. In this paper, we designARES, a method thatAutomaticallyReverseEngineers theSemantics of variables in PLC programs without requiring any domain knowledge.ARESis built on the fact that the Supervisory Control And Data Acquisition (SCADA) system monitors the behavior of PLC using a fixed mapping between the variables of program code and data log, and the data log variables are marked with physical semantics. By identifying the mapping between PLC code and SCADA data (i.e., the code-data mapping),ARESreverse engineers the physical semantics of program variables.ARESalso sheds light on the preferred defense strategies in implementing control rules that improve the resistance of PLC programs to semantic attacks, as well as in detecting and responding to semantics attacks in real time. We have experimentally evaluatedARESand the recommended defending practices on two ICS platforms.

Despite significantly enhancing system flexibility and reliability, the adoption of distributed secondary control in DC microgrids (DCmGs) introduces new vulnerabilities to false data injection (FDI) attacks. As a typical FDI attack, the zero trace stealthy (ZTS) attack has been recently disclosed for DCmGs, which can deteriorate the control objective while keeping stealthy to unknown input observer (UIO)-based detectors. In this work, we investigate the optimal deployment of ZTS attacks, where the adversary with limited resources aims to compromise a set of communication links such that the system state convergence error can be maximized. Specifically, we formulate the optimal ZTS attack deployment problem as a combinatorial optimization problem and unveil its NP-hard characteristic. Then, we discover the submodularity in the state convergence error function, enabling us to transform the original NP-hard problem into a tractable submodular maximization problem. Furthermore, based on the submodular optimization theory, we propose a novel distributed algorithm for the optimal ZTS attack deployment in DCmGs, which effectively balances the attack benefits and computation cost. Finally, comparisons between the centralized and distributed algorithms are illustrated through extensive simulations.

Industrial control systems (ICSs) govern the production activities of various critical infrastructures, where programmable logic controllers (PLCs) are essential devices for controlling industrial processes. However, PLCs have many vulnerabilities and might be configured inappropriately. With the trend of PLCs connecting to the Internet, such weaknesses will lead to various cyberattacks and have prompted many studies on the threat assessment for PLCs. Previous research has ignored PLCs’ security settings, such as operating mode and read/write authentication etc., which are the general security functionalities significantly affecting PLCs’ security. In this paper, we make the first attempt to propose a security settings-based threat assessment framework (SSTAF) to assess PLCs’ security.SSTAFconsists ofSScanner, a novel scanner to automatically extract the real-time configurations of security settings from PLCs, and the threat assessment criteria, serving to assess the appropriateness of PLC configurations and analyze risk levels of attacks based on PLCs’ security settings. Subsequently, usingSSTAF, we implement an Internet-wide threat assessment for PLCs exposed to the Internet. We deploySScanneron the Internet and interact with 41K ICS devices in cyberspace to acquire their configurations of security settings. Based on the scanning result and the threat assessment criteria, we reveal that 93.32% of PLCs have not appropriately configured their security settings. Additionally, each PLC might be subject to 4.96 attacks on average, of which 3.32 attacks are due to the inappropriate configurations of security settings.

The increasing network integration of industrial control systems amplifies the risk of cyberattacks on Programmable Logic Controllers (PLCs). In particular, the weak authentication of industrial communication protocols makes PLC program variables vulnerable to manipulation. Current defensive methods cannot reliably identify manipulated variables, even after PLC program manipulations have been detected. To bridge this gap, we presentADIS, a cross-domain Attack Detection and Identification System designed to detect and identify manipulated PLC program variables. Building on a novel state-aware graph representation of the PLC program,ADISdetects variable manipulations by comparing SCADA monitoring data with the control logic defined by the PLC program.ADISfurther identifies suspiciously manipulated program variables by excluding cascading failures from the detected anomalies and tracking suspicious variables based on the edges of the state-aware dependency graph. We have implemented and evaluatedADISon two platforms. The results demonstrate thatADISdetects attacks with a true positive rate exceeding 99% and a false positive rate of less than$0.04{\unicode {0x2030}}$. Furthermore, it successfully identifies manipulated program variables with up to a 71.3% reduction in suspicious variables compared to a baseline method.

Physical modules are the basic functional units of industrial control systems, governed by Programmable Logic Controllers (PLCs) according to predefined control logic. Attackers can compromise the integrity of physical modules by tampering with control logic, potentially disrupting production or causing physical damage. While model checking can detect logic bugs that violate module integrity requirements, it depends heavily on domain-specific knowledge, which is traditionally summarized by human experts, limiting both scalability and completeness. This paper proposes DGVerifier, a wiring diagram-guided framework that automatically verifies two general integrity requirements of physical modules: action integrity and state transition integrity. DGVerifier can extract module-related information from PLC wiring diagrams, mine domain-specific knowledge to generate specifications, and also model PLC programs as automata for verification. Evaluation on two real-world systems — an Elevator Control System and an Automated Assembly Line Control System — shows DGVerifier can recover 89.3% (25/28) of the required specifications and identify four hidden logic bugs violating physical module integrity.

With the rapid development of Connected and Autonomous Vehicles, In-Vehicle Network attacks have garnered heightened research scrutiny due to vehicles’ increasing connectivities to the external environment. The common characteristic among these attacks is to tamper with targeted powertrain-related signals in the Powertrain Controller Area Network (PT-CAN) and further physically threaten vehicles’ safety. These powertrain-related signals are encoded within CAN messages grounded by the syntax specification, which is proprietary to Original Equipment Manufacturers and publicly unavailable. Thus, to undertake comprehensive security analysis and strategies, reverse engineering PT-CAN to the semantic level is urgently needed. However, the existing methods rely on interactions (injecting challenge signals/actions) with the targeted vehicle, and certain manual efforts are required. To fill this gap, we proposePicaCAN, a novel framework to extract signals from CAN messages and reverse engineer their physical semantics based on physically induced causality. Once access to the CAN traffic,PicaCANoffers the researcher an eye on the vehicle’s powertrain system, decoding binaries flows into powertrain-related signals automatically. We experimentally evaluatePicaCANon PT-CAN of three automobiles containing two power types. The experimental results show thatPicaCANcould successfully extract physical signals representing all targeted semantics (pedals, engine speed, etc.) from two Internal Combustion Engine Vehicles and one Hybrid Electric Vehicle under EV mode.

TCP ensures reliable transmission through its stateful implementation and remains crucial today. TCP performance profiling is essential for tasks like diagnosing network performance problems, optimizing transmission performance, and developing new TCP variants, etc. Existing profiling methods lack enough attention to TCP state transition to provide detailed insights on TCP performance. Thus, we build TcpSight, a tool focusing on TCP state transition throughout connection lifetimes. TcpSight conducts stateful analysis by capturing key events using an efficient per-connection lock-free data management mechanism. Besides, TcpSight enhances profiling by integrating application layer information collected from the TCP stack. With the profiling results, users can identify the culprit of TCP performance degradation, and evaluate the performance of TCP algorithms. We design optional modules and filtering mechanisms to reduce TcpSights overhead. Our evaluation presents that TcpSight incurs an additional CPU consumption of about 16.6% (without filtering) and 10.6% (with filtering) when the servers load is 55.7%, and generates storage consumption about 1.88 KB per connection on average. We also give application cases of TcpSight and the deployment experiences in Alibaba Cloud. TcpSight helps in revealing meaningful findings and insights into exploiting TCP in the production deployment.

Symbolic execution is a powerful technique that can accurately synthesize program inputs for program testing. However, the scalability of symbolic execution is often limited by the capability of the constraint solver and time for testing. With limited time budget, it is desirable to optimally select paths for symbolic execution and furthermore variables for symbolization in order to achieve the maximum code coverage. In this work, we make two technical contributions towards solving this problem. First, different from most existing solving strategies based on heuristic path selection, we formally define the ‘optimal’ strategy based onthe reward of executing a given program path considering both possible code coverage and the cost of constraint solving. We further prove that the problem of identifying the optimal strategy for symbolic execution can be reduced to a classic knapsack problem, whose decision problem form is NP-complete. Second, in view of the complexity in identifying the optimal strategy, we design a practical greedy algorithm, named OPTSE, for approximating the optimal strategy. We implemented OPTSE in KLEE and extensively evaluate it on a diverse set of programs. The results show that OPTSE is effective, i.e., achieving 12% more code coverage and detects 17% more security violations than the state-of-the-art symbolic execution tool and outperforming a collection of strategies that only consider either path selection, solving strategies or simply superimpose them.

Recommender systems predict and suggest relevant options to users in various domains, such as e-commerce, streaming services, and social media. Recently, deep reinforcement learning (DRL)-based recommendation systems have become increasingly popular in academics and industry since DRL can characterize the long-term interaction between the system and users to achieve a better recommendation experience, e.g., Netflix, Spotify, Google, and YouTube.

Industrial protocols are widely used in Industrial Control Systems (ICSs) to network physical devices, thus playing a crucial role in securing ICSs. However, most commercial industrial protocols are proprietary and owned by their vendors, which impedes the implementation of protections against cyber threats. In this paper, we design REInPro to Reverse Engineer Industrial Protocols. REInPro is inspired by the fact that the structure of industrial protocols can be determined by a particular field referred to control field. By applying a probabilistic model of network traffic behavior, REInPro automatically identifies the control field and groups the associated network traffic into clusters. REInPro then infers critical semantics of industrial protocols by differentiating the features of corresponding protocol fields. We have experimentally implemented and evaluated REInPro using 8 different industrial protocols across 6 Programmable Logic Controllers (PLCs) belonging to 5 original equipment manufacturers. The experimental results show REInPro to reverse-engineer the formats and semantics of industrial protocols with an average correctness/perfection of 0.70/0.58 and 0.96/0.39.

Neural networks (NNs) are known to have diverse defects such as adversarial examples, backdoor and discrimination, raising great concerns about their reliability. While NN testing can effectively expose these defects to a significant degree, understanding their root causes within the network requires further examination. In this work, inspired by the idea of debugging in traditional software for failure isolation, we propose a novel unified neuron-isolation-based framework for debugging neural networks, shortly IDNN. Given a buggy NN that exhibits certain undesired properties (e.g., discrimination), the goal of IDNN is to identify the most critical and minimal set of neurons that are responsible for exhibiting these properties. Notably, such isolation is conducted with the objective that by simply ‘freezing’ these neurons, the model’s undesired properties can be eliminated, resulting in a much more efficient model repair compared to computationally expensive retraining or weight optimization as in existing literature. We conduct extensive experiments to evaluate IDNN across a diverse set of NN structures on five benchmark datasets, for solving three debugging tasks, including backdoor, unfairness, and weak class. As a lightweight framework, IDNN outperforms state-of-the-art baselines by successfully identifying and isolating a very small set of responsible neurons, demonstrating superior generalization performance across all tasks.

It is notoriously challenging to audit the potential unauthorized data usage in deep learning (DL) model development lifecycle, i.e., to judge whether certain private user data has been used to train or fine-tune a DL model without authorization. Yet, such data usage auditing is crucial to respond to the urgent requirements of trustworthy Artificial Intelligence (AI) such as data transparency, which are promoted and enforced in recent AI regulation rules or acts like General Data Protection Regulation (GDPR) and EU AI Act. In this work, we propose TeDA, a simple and flexible testing framework for auditing data usage in DL model development process. Given a set of user’s private data to protect (Dp), the intuition of TeDA is to apply membership inference (with good intention) for judging whether the model to audit (Ma) is likely to be trained with Dp. Notably, to significantly expose the usage under membership inference, TeDA applies imperceptible perturbation directed by boundary search to generate a carefully crafted test suite Dt (which we call ‘isotope’) based on Dp. With the test suite, TeDA then adopts membership inference combined with hypothesis testing to decide whether a user’s private data has been used to train Ma with statistical guarantee. We evaluated TeDA through extensive experiments on ranging data volumes across various model architectures for data-sensitive face recognition and medical diagnosis tasks. TeDA demonstrates high feasibility, effectiveness and robustness under various adaptive strategies (e.g., pruning and distillation).

Due to the vast testing space, the increasing demand for effective and efficient testing of deep neural networks (DNNs) has led to the development of various DNN test case prioritization techniques. However, the fact that DNNs can deliver high-confidence predictions for incorrectly predicted examples, known as the over-confidence problem, causes these methods to fail to reveal high-confidence errors. To address this limitation, in this work, we propose FAST, a method that boosts existing prioritization methods through guided FeAture SelecTion. FAST is based on the insight that certain features may introduce noise that affects the model's output confidence, thereby contributing to high-confidence errors. It quantifies the importance of each feature for the model's correct predictions, and then dynamically prunes the information from the noisy features during inference to derive a new probability vector for the uncertainty estimation. With the help of FAST, the high-confidence errors and correctly classified examples become more distinguishable, resulting in higher APFD (Average Percentage of Fault Detection) values for test prioritization, and higher generalization ability for model enhancement. We conduct extensive experiments to evaluate FAST across a diverse set of model structures on multiple benchmark datasets to validate the effectiveness, efficiency, and scalability of FAST compared to the state-of-the-art prioritization techniques.

The advancement of Industrial Internet-of-Things (IIoT) magnifies the cyber risk of automated production lines, especially to deception attacks that tamper with the monitoring data to prevent the manipulated operation of production lines from being detected. To address this issue, we propose Stochastic Manufacturing (StoM), a new paradigm of manufacturing that is resistant to deception by design. StoM voids the foundation of deception attacks — i.e., the highly predictable operation data due to the cyclical manufacturing process — by injecting controlled stochasticity into the operation of production lines without degrading manufacturing efficiency or quality. StoM then examines if this stochasticity can be observed from the operation data and triggers an alarm of deception attack if not. We have experimentally evaluated StoM on two production line platforms, showing StoM to detect deception attacks with a detection rate exceeding 99.1%, a false alarm rate below 0.1%, and a latency of less than 1.2 manufacturing cycles. Our empirical analysis also shows that it is highly impractical for attackers to spoof the controlled stochasticity.

We study learning adversarial Markov decision process (MDP) in the episodic setting under the constraint of differential privacy (DP). This is motivated by the widespread applications of reinforcement learning (RL) in non-stationary and even adversarial scenarios, where protecting users’ sensitive information is vital. We first propose two efficient frameworks for adversarial MDPs, spanning full-information and bandit settings. Within each framework, we consider both Joint DP (JDP), where a central agent is trusted to protect the sensitive data, and Local DP (LDP), where the information is protected directly on the user side. Then, we design novel privacy mechanisms to privatize the stochastic transition and adversarial losses. By instantiating such privacy mechanisms to satisfy JDP and LDP requirements, we obtain near-optimal regret guarantees for both frameworks. To our knowledge, these are the first algorithms to tackle the challenge of private learning in adversarial MDPs.

With the widespread application of collaborative learning (CL) technology in mobile-crowdsourcing-related scenarios, special attention should be paid to the privacy disclosure problem therein. Many pioneer noise-perturbation-based methods, particularly the differentially private ones, provide only homogeneous protection, which is insufficient for the heterogeneous protection requirements of many practical CL cases. In this article, we propose a privacy-aware mechanism that uses appropriate Gaussian noises to obfuscate the local and aggregated models. The noise variance is determined based on clients’ different privacy requirements. By zero-concentrated differential privacy, we analyze clients’ privacy-preserving degrees (PPDs) in the uplink and downlink channels. The obtained PPDs demonstrate that the information received by the aggregating server and the peer clients has distinct preservation effects, indicating that our scheme achieves the goal of heterogeneous protection. Moreover, we conduct a theoretical analysis of the performance of the global models aggregated during the iterative process. Finally, we validate the correctness of our theory with experimental results using a real-world data set.

Machine learning (ML) sees an increasing prevalence of being used in the internet-of-things (IoT)-based smart grid. However, the trustworthiness of ML is a severe issue that must be addressed to accommodate the trend of ML-based smart grid applications (MLsgAPPs). The adversarial distortion injected into the power signal will greatly affect the system’s normal control and operation. Therefore, it is imperative to conduct vulnerability assessment for MLsgAPPs applied in the safety-critical power systems. In this paper, we provide a comprehensive review of the recent progress in designing attack and defense methods for MLsgAPPs. Unlike the traditional survey about ML security, this is the first review work about the security of MLsgAPPs that focuses on the characteristics of power systems. We first highlight the specifics for constructing adversarial attacks on MLsgAPPs. Then, the vulnerability of MLsgAPP is analyzed from the perspective of the power system and ML model, respectively. Afterward, a comprehensive survey is conducted to review and compare existing studies about the adversarial attacks on MLsgAPPs in scenarios of generation, transmission, distribution, and consumption, and the countermeasures are reviewed according to the attacks that they defend against. Finally, the future research directions are discussed on the attacker’s and defender’s side, respectively. We also analyze the potential vulnerability of large language model-based (e.g., ChatGPT) smart grid applications. Overall, our purpose is to encourage more researchers to contribute to investigating the adversarial issues of MLsgAPPs.

Federated learning (FL) has emerged as a privacy-aware collaborative learning paradigm where participants jointly train a powerful model without sharing their private data. One desirable property for FL is the implementation of theright to be forgotten (RTBF), i.e., a leaving participant has the right to request the deletion of its private data from the global model. However,unlearning itself may not be enough to implement RTBF unless the unlearning effect can be independently verified, an important aspect that has been overlooked in the current literature. Unlearning verification is particularly challenging in FL as the unlearning effect on one participant's data could be canceled by the contribution of other participants. In this work, we prompt the concept ofverifiable federated unlearningand proposeVeriFi, a unified framework that allows systematic analysis of federated unlearning and quantification of its effect, with different combinations of various unlearning and verification methods. InVeriFi, the leaving participant is granted theright to verify (RTV)to actively verify the unlearning effect in the next few rounds immediately after notifying the server of its intention to leave, along with local verification done through two steps: 1)markingthat fingerprints the leaving participant by specially-designedmarkersand 2)checkingthat examines the global model's performance change on the markers. Based onVeriFi, we have conducted so far the most systematic study on verifiable federated unlearning, covering six unlearning methods and five verification methods. Our study sheds light on the existing drawbacks and potential alternatives for both unlearning and verification methods. During the study, we also propose a more efficient and FL-friendly unlearning method$^{u}$S2U, and two more effective and robust non-invasive (without training controllability, external data, white-box model access nor introducing new security risks) verification methods$^{v}$FM and$^{v}$EM. While the proposed methods may not be a panacea for all the challenges, they address several key drawbacks of existing methods and represent a promising step toward effective, efficient, robust, and more importantly, non-invasive federated unlearning and verification. We extensively evaluateVeriFion seven datasets, including natural/facial/medical images and audios, and four types of deep learning models, including both Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). We hope, such an extensive and holistic experimental evaluation, although admittedly complex and challenging, could help establish important empirical understandings, evidence, and insights for trustworthy federated unlearning.

The widespread use of programmable logic controllers (PLCs) in critical infrastructures has given rise to escalating cybersecurity concerns regarding PLC attacks. As a proactive defense mechanism, PLC honeypots emulate genuine controllers to engage adversaries so as to observe their attack tactics and techniques. As part of the arms race between the offense and defense, multiple PLC honeypot identification tools have been developed. However, many existing tools cannot recognize high-fidelity honeypots, since they rely on identifying common network services and fingerprints. In this paper, we propose an innovative and practical honeypot identification framework calledHoneyJudge, which goes beyond state-of-the-art (SOTA) network fingerprint-based identification tools like Nmap and the PLCScan tool.HoneyJudgetests the suspected target’s special memory content and features. Specifically,HoneyJudgemodels the internal memory of a PLC in three categories, from system-level, user-level, to process-level categories, based on which it extracts six representative memory features. All characteristics are acquired through automated network request messages. Then, we design a weighted voting algorithm to combine the test results over different memory features to reach the final conclusion. We validate the effectiveness ofHoneyJudgein comparison with several SOTA honeypot identification tools, and the results indicate that the memory-related issues have not been well addressed in existing PLC honeypots and still need substantial research efforts.

The information regarding the driving states of all vehicles is crucial for achieving optimal group performance in a vehicle platoon. This article focuses on the fully distributed driving state estimation problem in open vehicle platoons, which frequently experience arrivals and departures of vehicles. To address this problem, we propose a distributed driving state observer inspired by the leader–follower consensus technique. This observer can reconstruct the global driving state of the platoon, including the positions, velocities, and accelerations of all vehicles. We also derive the necessary and sufficient conditions to ensure the stability of its estimation error dynamics. The proposed observer is highly flexible in platoons with a strongly connected communication network, as it can be constructed and operated using the local knowledge of each vehicle only, without relying on global information of a platoon such as the number of vehicles. We demonstrate the observer's plug-and-play operations in the face of platoon merging and splitting and analyze its estimation stability. Extensive simulation results demonstrate the effectiveness of our theoretical results and the potential of the proposed observer for platoon control.

Vertical federated learning (VFL) revolutionizes privacy-preserved collaboration for small businesses that have distinct but complementary feature sets. However, as the scope of VFL expands, the constant entering and leaving of participants and the subsequent exercise of the “right to be forgotten” pose a great challenge in practice. The question of how to efficiently erase one’s contribution from the shared model remains largely unexplored in the context of VFL. In this article, we introduce a vertical federated unlearning framework, which integrates model checkpointing techniques with a hybrid, first-order optimization technique. The core concept is to reduce backpropagation time and improve convergence/generalization by combining the advantages of the existing optimizers. We provide in-depth theoretical analysis and time complexity to illustrate the effectiveness of the proposed design. We conduct extensive experiments on six public datasets and demonstrate that our method could achieve up to 6.3× speedup compared to the baseline, with negligible influence on the original learning task.

The Virtual Private Cloud (VPC) service enables users to configure shared resources within public clouds on demand, providing isolation between users. However, configuring the VPC network is a complex and error-prone task, and misconfiguration has been the leading cause of cloud network security issues. The large number of complex network components and configurations makes it difficult to perform scalable, efficient, and accurate fault verification of the network behavior. To address this issue, we design a comprehensive and automated fault diagnosis and localization tool, calledLFVeri, which is built upon an innovative modular network model that accurately captures the logic functions of real components within VPC networks, and propose eleven functions to verify network reachability and security requirements. We conduct performance testing ofLFVerion various datasets and compared it with other verification tools. The experiments show thatLFVerioutperforms in modeling and analyzing real VPC scenarios while also possessing the fastest verification speed. It can model and analyze large VPC networks with tens of thousands of components and millions of configuration rules in less than half an hour.

Fuzzing is one of the prevailing methods for vulnerability detection. However, even state-of-the-art fuzzing methods become ineffective after some period of time, i.e., the coverage hardly improves as existing methods are ineffective to focus the attention of fuzzing on covering the hard-to-trigger program paths. In other words, they cannot generate inputs that can break the bottleneck due to the fundamental difficulty in capturing the complex relations between the test inputs and program coverage. In particular, existing fuzzers suffer from the following main limitations: 1) lacking an overall analysis of the program to identify the most “rewarding” seeds, and 2) lacking an effective mutation strategy which could continuously select and mutates the more relevant “bytes” of the seeds. In this work, we propose an approach calledATTuzzto address these two issues systematically. First, we propose a lightweight dynamic analysis technique that estimates the “reward” of covering each basic block and selects the most rewarding seeds accordingly. Second, we mutate the selected seeds according to a neural network model which predicts whether a certain “rewarding” block will be covered given certain mutations on certain bytes of a seed. The model is a deep learning model equipped with an attention mechanism which is learned and updated periodically whilst fuzzing. Our evaluation shows thatATTuzzsignificantly outperforms 5 state-of-the-art grey-box fuzzers on 6 popular real-world programs and MAGMA data sets at achieving higher edge coverage and finding new bugs. In particular,ATTuzzachieved 1.2X edge coverage and 1.8X bugs detected than AFL++ over 24-hour runs. In addition,ATTuzzalso finds 4 new bugs in the latest version of some popular software including p7zip and openUSD.

Programmable logic controllers (PLCs), the essential components of critical infrastructure, play a crucial role in various industrial manufacturing processes. Recent attack events show that attackers have a strong interest in tampering with the controller variables, such as the device status and internal program logic. A typical attack strategy is that the attackers just send malicious network traffic of industrial control protocols (ICPs) to change the controller variables of PLCs. To defend against this attack, a lot of countermeasures have been proposed to detect anomalies in network traffic based on the semantic analysis.

The widespread deployment of storage disaggregation in the cloud has facilitated flexible scaling and storage overprovisioning, allowing for high utilization of storage capacity and IOPS. Instead of utilizing remote storage protocols to access remote disks, a middle-tier is introduced between compute servers and storage servers in order to serve I/O requests from compute servers and provide computations such as compression and decompression. However, due to the need for a cloud to concurrently serve millions of VMs that require access to disaggregated storage, the middle-tier requires a massive number of servers to process network traffic between computing and storage nodes. For example, a major cloud company may deploy hundreds of thousands of high-end servers to provide such a service for its cloud storage, because the existing CPU-based middle-tier suffers from a severe issue of compute-intensive compression/decompression on high-throughput storage traffic. To address this issue, we introduce SmartDS, a middle-tier-centric SmartNIC that serves storage I/O requests with low latency and high throughput, while maintaining high flexibility and programmability. The key idea behind SmartDS is the application-aware message split (AAMS) mechanism, which allows for the processing of the message's header on the host CPU to achieve high flexibility, and the message's payload on the SmartDS. Experimental results demonstrate that SmartDS provides up to 4.3× more throughput than a CPU-based middle-tier and enables the linear scale-up of multiple network ports and multiple SmartNICs, thus significantly reducing cloud infrastructure costs for disaggregated block storage.

Cyber–physical systems (CPSs) are ubiquitous in critical infrastructures, where programmable logic controllers (PLCs) and physical components intertwine. However, multiple successful attacks targeting safety-related CPSs, in particular the PLCs, manifest their vulnerability toward malicious cyber attacks, which may cause significant damage consequently. Though several kinds of defending techniques exist in the literature, few of them can be practically and widely applied to real-world CPSs equipped with PLCs from leading vendors, primarily due to the lack of specific hardware or unrealistic defense assumptions. In this article, we propose PLC-READER, a practical memory attacks detection and response framework to secure the CPS. The core of PLC-READER includes: 1) a comprehensive semantic analysis solution specifically for PLC’s proprietary protocol based on software reverse engineering and network traffic difference analysis and 2) a fine-grained memory structure analysis solution to identify the critical memory data. Based on the results of such reverse engineering, PLC-READER further performs sanity checks for the PLC’s critical memory by periodically checking the hash values and dynamic checksum values of these memory data. We extensively evaluated PLC-READER against four types of 366 different memory attacks, with some newly developed ones which got six CVE IDs from Schneider and Rockwell, by analyzing three kinds of proprietary protocols and six kinds of memory structures in six kinds of real-world PLCs from three leading manufacturers. The results demonstrate that the PLC-READER can detect all memory attacks with an accuracy of 100% and perform corresponding emergency responses in time.

Distributed classification learning (DCL) is a promising solution to establish Internet of Things-based smart applications, especially due to its strong ability in dealing with large-scale and high-concurrency data. However, the performance of DCL may be seriously affected by the label flipping attack (LFA). Regarding the LFA-resilient learning problem, most existing works are built in more centralized settings. The work addressing the secure DCL issue makes an assumption that the label flipping rates are symmetric and available for scheme design. In this article, we remove this assumption and propose an LFA-resilient DCL scheme, named FENDER, without knowing the asymmetric flipping rates. The challenge is to guarantee both attack resilience and algorithm convergence. We carefully integrate a resilient loss and the alternating direction method of the multiplier scheme, making FENDER resilient to LFA. Further, we systematically analyze the performance of FENDER according to a metric reflecting the models obtained by all the servers at different iterations. In addition, we discuss and compare FENDER with some existing methods from the aspects of algorithm establishment and performance guarantee. Finally, extensive experiments with multiple real-world data sets are performed to validate the developed theory and evaluate the performance of the trained models.

For public cloud providers, it is of great significance to maintain the availability of their cloud services, which requires efficient anomaly diagnosis and recovery. To achieve such properties, the first step is to localize the anomalies, i.e., determining where they happen in the network path of cloud-client services. We propose FlowPinpoint to perform anomaly localization for cloud providers. FlowPinpoint collects statistics of each network flow at the cloud network gateways (i.e., gateway flowlog), where the collected data can reflect the information from both the cloud side and the Internet side. Aggregation and association are conducted on the datacenter-scale gateway flowlogs by Alibaba's big data computing platform. In order to preclude the disturbance of anomaly-unrelated flowlogs, a two-layer filter is proposed which consists of an indicator-based filter and an isolation forest filter. Finally, the anomaly localization analyzer classifies the flowlogs and determines whether the anomaly is inside the cloud network or not according to the classification results. FlowPinpoint is implemented and tested in the production environment of Alibaba Cloud, and it correctly localizes 1 anomaly inside the cloud and 6 anomalies on the Internet over 4 months.

As a critical infrastructure, the traditional power system has transformed into a cyber-physical integrated smart grid. However, the vulnerabilities exposed in either the cyber or physical layer might be exploited by adversaries to construct complicated and coordinated attacks consequent in destructive impacts. In this paper, we propose a stealthy physics-manipulated attack (SPMA) by masking the physical attacks on the flexible AC transmission system (FACTS) with strategic cyberattacks. To construct the SPMA, we first manipulate the control command sent to the FACTS device to change the reactance and then tamper with the sensor measurements to conceal it. The SPMA is constructed with complete-informed and incomplete-informed attackers, noisy sensor measurements, and a nonlinear AC model, respectively. The impact of the physics manipulation on the real-time economic dispatch and the system’s operation security are formulated and numerically analyzed. Furthermore, we also provide potential countermeasures from three aspects to defend against SPMAs. Finally, extensive experiments are conducted with the IEEE test power systems to evaluate the stealthiness of SPMAs and the economic losses and potential cascading failures caused by SPMAs using real-world load profiles.

The watermarking method, which adds unique watermarks to data, has been widely used for integrity attack detection in industrial control systems (ICSs). Existing literature generally designs watermarking mechanisms without considering the existence of noises, which cannot be trivially applied to realistic ICS scenarios in the presence of strong noise interference. On one hand, the low-intensity watermarking will be ineffective under the strong noise environment; while on the other hand, the oversized watermarking can possibly degrade the control performance or even destabilize the system. Therefore, the intensity of watermarks plays a fundamental role in balancing the tradeoff between detection effectiveness and control performance, which, to the best of our knowledge, has never been thoroughly analyzed yet. To this end, in this paper, we for the first time propose an optimal watermarking design method for ICSs considering the detection-performance tradeoff. To begin with, we shift the watermark container from data points to segments and update the detection metrics to reduce the noise impact. Then, we formulate an optimization problem to determine the strength of watermarks to balance the detection-performance tradeoff. Meanwhile, the detection effectiveness and control performance metrics are analytically modeled and theoretically analyzed considering the discrepancy between added watermarks and noises, signal quality, detection latency, as well as estimation of detection metrics. Finally, extensive numerical simulations and systematical experiments based on a practical Ethanol Distillation ICS are conducted to validate the theoretical analysis and demonstrate the outperformance of our proposed watermarking method in comparison with related works.

Recently, there has been significant growth of interest in applying software engineering techniques for the quality assurance of deep learning (DL) systems. One popular direction is DL testing—that is, given a property of test, defects of DL systems are found either by fuzzing or guided search with the help of certain testing metrics. However, recent studies have revealed that the neuron coverage metrics, which are commonly used by most existing DL testing approaches, are not necessarily correlated with model quality (e.g., robustness, the most studied model property), and are also not an effective measurement on the confidence of the model quality after testing. In this work, we address this gap by proposing a novel testing framework called QuoTe (i.e., Qu ality- o riented Te sting). A key part of QuoTe is a quantitative measurement on (1) the value of each test case in enhancing the model property of interest (often via retraining) and (2) the convergence quality of the model property improvement. QuoTe utilizes the proposed metric to automatically select or generate valuable test cases for improving model quality. The proposed metric is also a lightweight yet strong indicator of how well the improvement converged. Extensive experiments on both image and tabular datasets with a variety of model architectures confirm the effectiveness and efficiency of QuoTe in improving DL model quality—that is, robustness and fairness. As a generic quality-oriented testing framework, future adaptations can be made to other domains (e.g., text) as well as other model properties.

Programmable Logic Controllers (PLCs) are responsible for automating process control in many industrial systems (e.g. in manufacturing and public infrastructure), and thus it is critical to ensure that they operate correctly and safely. The majority of PLCs are programmed in languages such as Structured Text (ST). However, a lack of formal semantics makes it difficult to ascertain the correctness of their translators and compilers, which vary from vendor-to-vendor. In this work, we develop K-ST, a formal executable semantics for ST in the$\boldsymbol{\mathbb{K}}$framework. Defined with respect to the IEC 61131-3 standard and PLC vendor manuals, K-ST is a high-level reference semantics that can be used to evaluate the correctness and consistency of different ST implementations. We validate K-ST by executing 567 ST programs extracted from GitHub and comparing the results against existing commercial compilers (i.e., CODESYS, CX-Programmer, and GX Works2). We then apply K-ST to validate the implementation of the open source OpenPLC platform, comparing the executions of several test programs to uncover five bugs and nine functional defects in the compiler.

Moving target defense (MTD) is a new defensive mechanism developed in power systems to thwart false data injection attacks (FDIAs). However, since the MTD works by perturbing the branch parameters with the distributed flexible ac transmission system (D-FACTS), it might cause additional infrastructure and operation costs and affect the system dynamics. This is a complicated problem because it is closely related to which branches should be perturbed and how much they are changed. In this article, we analyze the essentials of MTD and construct an effective and low-cost MTD. To begin with, we provide a sufficient and necessary condition for MTD to protect a bus from being affected by the intended FDIA. Based on this result, we propose a new metric to quantify the protection level of MTD and an efficient algorithm to minimize the number of required D-FACTS devices for protecting a specific set of buses. To reduce the operation cost, we develop two strategies to make the increasing operation cost zero for activating the MTD. Furthermore, we analyze the impact of MTD on the system dynamics with a special emphasize on small signal stability. Finally, we conduct extensive simulations to validate our findings with the test cases of power systems in MATPOWER.

Neural backdoors represent one primary threat to the security of deep learning systems. The intensive research has produced a plethora of backdoor attacks/defenses, resulting in a constant arms race. However, due to the lack of evaluation benchmarks, many critical questions remain under-explored: (i) what are the strengths and limitations of different attacks/defenses? (ii) what are the best practices to operate them? and (iii) how can the existing attacks/defenses be further improved? To bridge this gap, we design and implement TROJAN-ZOO, the first open-source platform for evaluating neural backdoor attacks/defenses in a unified, holistic, and practical manner. Thus far, focusing on the computer vision domain, it has incorporated 8 representative attacks, 14 state-of-the-art defenses, 6 attack performance metrics, 10 defense utility metrics, as well as rich tools for in-depth analysis of the attack-defense interactions. Leveraging TROJANZOO, we conduct a systematic study on the existing attacks/defenses, unveiling their complex design spectrum: both manifest intricate trade-offs among multiple desiderata (e.g., the effectiveness, evasiveness, and transferability of attacks). We further explore improving the existing attacks/defenses, leading to a number of interesting findings: (i) one-pixel triggers often suffice; (ii) training from scratch often outperforms perturbing benign models to craft trojan models; (iii) optimizing triggers and trojan models jointly greatly improves both attack effectiveness and evasiveness; (iv) individual defenses can often be evaded by adaptive attacks; and (v) exploiting model interpretability significantly improves defense robustness. We envision that TROJANZOO will serve as a valuable platform to facilitate future research on neural backdoors.

Semantic attacks have incurred increasing threats to Industrial Control Systems (ICSs), which manipulate targeted system modules by identifying the physical semantics of variables in Programmable Logic Controllers (PLCs) programs, i.e., the sensing/actuating modules represented by the variables. This is usually (and inefficiently) achieved via manual examination of system documents and long-term observation of system behavior. In this paper, we design ARES, a method that Automatically Reverse Engineers the Semantics of variables in PLC programs without requiring any domain knowledge. ARES is built on the fact that the Supervisory Control And Data Acquisition (SCADA) system monitors the behavior of PLC using a fixed mapping between the variables of program code and data log, and the data log variables are marked with physical semantics. By identifying the mapping between PLC code and SCADA data (i.e., the code-data mapping), ARES reverse engineers the physical semantics of program variables. ARES also sheds light on the preferred practices in implementing control rules that improve the resistance of PLC programs to semantic attacks. We have experimentally evaluated ARES and the recommended implementation practices on two ICS platforms.

Deep learning models, especially those large-scale and high-performance ones, can be very costly to train, demanding a considerable amount of data and computational resources. As a result, deep learning models have become one of the most valuable assets in modern artificial intelligence. Unauthorized duplication or reproduction of deep learning models can lead to copyright infringement and cause huge economic losses to model owners, calling for effective copyright protection techniques. Existing protection techniques are mostly based on watermarking, which embeds an owner-specified watermark into the model. While being able to provide exact ownership verification, these techniques are 1) invasive, i.e., they need to tamper with the training process, which may affect the model utility or introduce new security risks into the model; 2) prone to adaptive attacks that attempt to remove/replace the watermark or adversarially block the retrieval of the watermark; and 3) not robust to the emerging model extraction attacks. Latest fingerprinting work on deep learning models, though being non-invasive, also falls short when facing the diverse and ever-growing attack scenarios.In this paper, we propose a novel testing framework for deep learning copyright protection: DEEPJUDGE. DEEPJUDGE quantitatively tests the similarities between two deep learning models: a victim model and a suspect model. It leverages a diverse set of testing metrics and efficient test case generation algorithms to produce a chain of supporting evidence to help determine whether a suspect model is a copy of the victim model. Advantages of DEEPJUDGE include: 1) non-invasive, as it works directly on the model and does not tamper with the training process; 2) efficient, as it only needs a small set of seed test cases and a quick scan of the two models; 3) flexible, i.e., it can easily incorporate new testing metrics or test case generation methods to obtain more confident and robust judgement; and 4) fairly robust to model extraction attacks and adaptive attacks. We verify the effectiveness of DEEPJUDGE under three typical copyright infringement scenarios, including model finetuning, pruning and extraction, via extensive experiments on both image classification and speech recognition datasets with a variety of model architectures.

Programmable logic controllers (PLCs), i.e., the core of control systems, are well-known to be vulnerable to a variety of cyber attacks. To mitigate this issue, we designPLC-Sleuth, a novel noninvasive intrusion detection/localization system for PLCs, which is built on a set of control invariants—i.e., the correlations between sensor readings and the concomitantly triggered PLC commands—that exist pervasively in all control systems. Specifically, taking the system’s supervisory control and data acquisition log as input,PLC-Sleuthabstracts/identifies the system’s control invariants as a control graph using data-driven structure learning, and then monitors the weights of graph edges to detect anomalies thereof, which is in turn, a sign of intrusion. We have implemented and evaluatedPLC-Sleuthusing both a platform of ethanol distillation system (EDS) and a realistically simulated Tennessee Eastman (TE) process. The results show thatPLC-Sleuthcan: 1) identify control invariants with 100%/98.11% accuracy for EDS/TE; 2) detect PLC intrusions with 98.33%/0.85 ‰ true/false positives (TPs/FPs) for EDS and 100%/0% TP/FP for TE; and 3) localize intrusions with 93.22%/96.76% accuracy for EDS/TE.

With the widespread adoption of Internet of Things (IoT) technologies, cyber–physical systems (CPSs) are facing threats from cyberattacks due to the vulnerabilities exposed in IoT devices. In this article, we analyze the feasibility of a coordinated attack, named TD-FDIA, on CPS by the synchronizing the time-delay attack (TDA) and false data injection attack (FDIA). It seems that the coordinated attack is more powerful than either one. But the analysis of its stealthiness and effectiveness is challenging. In the context of the networked control system, we first propose a general formulation for the impact of TD-FDIA on the system’s stability. Then, we analyze whether the combination of TDA and FDIA can destabilize the system and remain stealthy or not with different setups when the controller is with and without an observer, and the communication protocol between the controller and actuator is UDP and TCP, respectively. The conditions required to make TD-FDIA stealthy are given in some cases. Finally, we conduct extensive experiments to evaluate the impact of TDA, FDIA, and TD-FDIA on the system’s stability with three different CPS scenarios.

Deploying machine learning (ML)-based intrusion detection systems (IDS) is an effective way to improve the security of industrial control systems (ICS). However, ML models themselves are vulnerable to adversarial examples, generated by deliberately adding subtle perturbation to the input sample that some people are not aware of, causing the model to give a false output with high confidence. In this article, our goal is to investigate the possibility of stealthy cyber attacks towards IDS, including injection attack, function code attack and reconnaissance attack, and enhance its robustness to adversarial attack. However, adversarial algorithms are subject to communication protocol and legal range of data in ICS, unlike only limited by the distance between original samples and newly generated samples in image domain. We propose two strategies - optimal solution attack and GAN attack - oriented to flexibility and volume of data, formulating an optimization problem to find stealthy attacks, where the former is appropriate for not too large and more flexible samples while the latter provides a more efficient solution for larger and not too flexible samples. Finally, we conduct experiments on a semi-physical ICS testbed with a high detection performance ensemble ML-based detector to show the effectiveness of our attacks. The results indicate that new samples of reconnaissance and function code attack produced by both optimal solution and GAN algorithm possess 80 percent higher probability to evade the detector, still maintaining the same attack effect. In the meantime, we adopt adversarial training as a method to defend against adversarial attack. After training on the mixture of orginal dataset and newly generated samples, the detector becomes more robust to adversarial examples.

Industrial robots are prototypical cyber-physical systems widely deployed in (smart) manufacturing, which operate according to the operation code uploaded by the human operator and are monitored in real-time based on their movement data. However, industrial robots suffer from replay attacks, via which attackers can manipulate the robot operation without being observed by the monitoring system. To mitigate this vulnerability, we design a novel intrusion detection system for industrial robots using their power fingerprint, calledPIDS(Power-basedIntrusionDetectionSystem), and deliverPIDSas abump-in-the-wiremodule installed at the powerline of commodity robots. The foundation ofPIDSis the physically-induced dependency between the robot movement and the concomitant power consumption, whichPIDScaptures via joint physical analysis and (cyber) data-driven modeling.PIDSthen fingerprints the robot movements observed by the monitoring system using their expected power consumption, and cross-validates the fingerprints with empirically collected power information — a mismatch thereof flags anomalies of the observed movements (i.e., evidence of replay attack). We have evaluatedPIDSusing three models of robots from different vendors — i.e., ABB IRB120, KUKA KR6 R700, and Universal Robots UR5 robots — with over 2,000 operation cycles. Experimental results show thatPIDSdetects replay attacks at an average rate of 96.5 percent (up to 99.9 percent) and a 0.1s latency.

Persistent packet loss in the cloud-scale overlay network severely compromises tenant experiences. Cloud providers are keen to diagnose such problems efficiently. However, existing work is either designed for the physical network or insufficient to present the concrete reason of packet loss. We propose to record and analyze the on-site forwarding condition of packets during packet-level tracing. The cloud-scale overlay network presents great challenges to achieve this goal with its high network complexity, multi-tenant nature, and diversity of root causes. To address these challenges, we present VTrace, an automatic diagnostic system for persistent packet loss over the cloud-scale overlay network. Utilizing the “fast path-slow path” structure of virtual forwarding devices (VFDs), e.g., vSwitches, VTrace installs several “coloring-matching-logging” rules in VFDs to selectively track the target packets and inspect them in depth. The detailed forwarding situation at each hop is logged and then assembled to perform analysis with an efficient path reconstruction scheme. Experiments are conducted to demonstrate VTrace’s low overhead and quick response. Besides, based on the idea “coloring-matching-counting”, VTrace can be easily extended toVTrace-statsto identify the culprit device for transient packet loss. We share experiences of how VTrace andVTrace-statsefficiently work after deploying them in Alibaba Cloud for years.

For protecting users' private data, local differential privacy (LDP) has been leveraged to provide the privacy-preserving range query, thus supporting further statistical analysis. However, existing LDP-based range query approaches are limited by their properties, ie, collecting user data according to a pre-defined structure. These static frameworks would incur excessive noise added to the aggregated data especially in the low privacy budget setting. In this work, we propose an Adaptive Hierarchical Decomposition (AHEAD) protocol, which adaptively and dynamically controls the built tree structure, so that the injected noise is well controlled for maintaining high utility. Furthermore, we derive a guideline for properly choosing parameters for AHEAD so that the overall utility can be consistently competitive while rigorously satisfying LDP. Leveraging multiple real and synthetic datasets, we extensively show the effectiveness of AHEAD in both low and high dimensional range query scenarios, as well as its advantages over the state-of-the-art methods. In addition, we provide a series of useful observations for deploying \myahead in practice.

Recently, there has been a significant growth of interest in applying software engineering techniques for the quality assurance of deep learning (DL) systems. One popular direction is deep learning testing, where adversarial examples (a.k.a.~bugs) of DL systems are found either by fuzzing or guided search with the help of certain testing metrics. However, recent studies have revealed that the commonly used neuron coverage metrics by existing DL testing approaches are not correlated to model robustness. It is also not an effective measurement on the confidence of the model robustness after testing. In this work, we address this gap by proposing a novel testing framework called Robustness-Oriented Testing (RobOT). A key part of RobOT is a quantitative measurement on 1) the value of each test case in improving model robustness (often via retraining), and 2) the convergence quality of the model robustness improvement. RobOT utilizes the proposed metric to automatically generate test cases valuable for improving model robustness. The proposed metric is also a strong indicator on how well robustness improvement has converged through testing. Experiments on multiple benchmark datasets confirm the effectiveness and efficiency of RobOT in improving DL model robustness, with 67.02% increase on the adversarial robustness that is 50.65% higher than the state-of-the-art work DeepGini.

IoT devices are abnormally prone to diverse errors due to harsh environments and limited computational capabilities. As a result, correct error handling is critical in IoT. Implementing correct error handling is non-trivial, thus requiring extensive testing such as fuzzing. However, existing fuzzing cannot effectively test IoT error-handling code. First, errors typically represent corner cases, thus are hard to trigger. Second, testing error-handling code would frequently crash the execution, which prevents fuzzing from testing following deep error paths.In this paper, we propose IFIZZ, a new bug detection system specifically designed for testing error-handling code in Linux-based IoT firmware. IFIZZ first employs an automated binary-based approach to identify realistic runtime errors by analyzing errors and error conditions in closed-source IoT firmware. Then, IFIZZ employs state-aware and bounded error generation to reach deep error paths effectively. We implement and evaluate IFIZZ on 10 popular IoT firmware. The results show that IFIZZ can find many bugs hidden in deep error paths. Specifically, IFIZZ finds 109 critical bugs, 63 of which are even in widely used IoT libraries. IFIZZ also features high code coverage and efficiency, and covers 67.3% more error paths than normal execution. Meanwhile, the depth of error handling covered by IFIZZ is 7.3 times deeper than that covered by the state-of-the-art method. Furthermore, IFIZZ has been practically adopted and deployed in a worldwide leading IoT company. We will open-source IFIZZ to facilitate further research in this area.

Data integrity attack (DIA) is one class of threatening cyber attacks against the Internet-of-Things (IoT)-based smart grid. With the assumption that the attacker is capable of obtaining complete or incomplete information of the system topology and branch parameters, it has been widely recognized that the highly synthesized DIA can evade being detected and undermine the smart grid state estimation. However, the branch parameters cannot be easily obtained or inferred by the attacker in practice. They can be changed or disturbed with time. In this article, we complete the class of DIA by designing the zero-parameter-information DIA (ZDIA), which makes it possible for the attacker to execute stealthy data tampering attacks without any information of the branch parameters. Only the topology information about the cut line is required to construct such attack. We prove that, the attacker can arbitrarily modify the state estimate of a one-degree bus, which is connected to the outside only by a single cut line; and modify the state estimates of all buses, with the same arbitrary bias, in a one-degree super-bus, which is a group of buses that is connected to the outside only by a single cut line. Besides, we extend ZDIA to the cases where a bus and super-bus are connected to the outside only by several cut lines. Moreover, we propose two countermeasures to address the topology vulnerability exploited by ZDIA, and present a branch perturbation strategy to defend against general DIAs. Finally, we conduct extensive simulations with the IEEE standard power systems to validate the theoretical results.

Control performance of vehicle platooning relies on the information flow topology and quality of wireless communications. In this article, we investigate the constant-time-headway-spacing-policy-based vehicle platooning problem, where multiple predecessors' information is used by the following vehicles and communication impairments, i.e., limited communication range and random packet losses, are considered. In this article, first, when the leading vehicle moves at a constant speed, we obtain the sufficient and necessary conditions on sampling time, control gains, and internal lag, to ensure the stability of the vehicle platoon based on matrix polynomials' stability for ideal communications. Second, for time-independent homogeneous random packet losses, we provide the upper bound for the loss rate to maintain convergence in expectation by matrix eigenvalue perturbation theory when no input is set for lossy information. We also provide sufficient conditions to guarantee mean-square convergence for heterogeneous time-independent random packet losses and show the convergence time for any given accuracy and probability. Third, when historically latest information is used for input, the sufficient and necessary conditions are provided to ensure the internal stability and string stability by Markov jump linear system theory. Furthermore, we discuss the controller design when no feasible solution exists to guarantee the string stability. Extensive numerical results validate our analysis.

Recently, cloud computing has become a vital part that supports people’s normal lives and production. However, accompanied by the increasing complexity of the cloud network, failures constantly keep coming up and cause huge economic losses. Thus, to guarantee the cloud network performance and prevent execrable effects caused by failures, cloud network diagnostics has become of great interest for cloud service providers. Due to the characteristics of cloud network (e.g., virtualization and multi-tenancy), transplanting traditional network diagnostic tools to the cloud network face several difficulties. Additionally, many existing tools cannot solve problems in the cloud network. In this paper, we summarize and classify the state-of-the-art technologies of cloud diagnostics which can be used in the production cloud network according to their features. Moreover, we analyze the differences between cloud network diagnostics and traditional network diagnostics based on the characteristics of the cloud network. Considering the operation requirements of the cloud network, we propose the points that should be cared about when designing a cloud network diagnostic tool. Also, we discuss the challenges that cloud network diagnostics will face in future development.

Collaborative computing is efficient to conduct large-scale computation tasks, especially with the surge in data volume. However, when the data contains sensitive information, privacy has to be attached significant attention during the execution of computation tasks. In this paper, based on a two-step average computation framework, we first propose three different privacy-aware schemes, where noises are carefully designed to be injected into the distributed computing process. The challenging issue is to guarantee the privacy loss in each iteration to be controllable and quantifiable, which we call the dynamic privacy-preserving collaborative computing problem. By employing Kullback-Leibler differential privacy, we obtain the privacy preserving levels in different iterations regarding the three schemes, followed by the analysis of their convergence performances. Further, we devise an approach to balance the privacy loss and the computation accuracy, whose challenge lies in how to motivate data contributors (DCs) to report more accurate data without providing them with monetized payments. This is done by allowing DCs to report data multiple times, and we obtain the optimal reporting times for each DC. Finally, extensive numerical experiments are performed to validate the obtained theoretical results.

Core devices form the critical components of the cloud network and provide service to multiple tenants simultaneously. The anomalies that happened in core devices impact network availability of a large number of users, meanwhile, lead to the degradation of cloud providers’ profits. However, direct monitoring of core devices needs to deploy massive heartbeat checking tools on numerous related components, which will be extremely laborious. In this paper, we deploy RAIN to reduce the number of devices that need to be detailed investigated for anomalies. The session traffic data among core devices and served virtual machines are utilized to conduct the analyzing. To guarantee near real-time monitoring, RAIN is designed as a two-step structure and incorporating four feature-based detection methods. RAIN has been deployed in Alibaba’s production cloud network for over 6 months and is analyzing terabytes of traffic flow metrics per day.

Industrial robots have been shown to suffer from replay attacks, via which adversaries not only manipulate the robot operation by downloading malicious code, but also prevent the detection of this manipulation by replaying recorded (and normal) movement data to the monitoring system. To protect industrial robots from replay attacks, we design a novel intrusion detection system using the power fingerprint of robots, called PIDS (Power-based Intrusion Detection System), and deliver PIDS as a bump-in-the-wire module installed at the powerline of commodity robots. The foundation of PIDS is the physically-induced dependency between the robot movement and the concomitant electrical power consumption, which PIDS captures via joint physical analysis and (cyber) data-driven modeling. PIDS then fingerprints the robot movements observed by the monitoring system using their expected power consumption, and cross-validates the fingerprints with empirically collected power information --- a mismatch thereof flags anomalies of the observed movements (i.e., evidence of replay attack). We have evaluated PIDS using three models of robots from different vendors --- i.e., ABB IRB120, KUKA KR6 R700, and Universal Robots UR5 robots --- with over 2, 000 operation cycles. The experimental results show that PIDS detects replay attacks with an average rate of 96.5% (up to 99.9%) and a 0.1s latency.

Persistent packet loss in the cloud-scale overlay network severely compromises tenant experiences. Cloud providers are keen to automatically and quickly determine the root cause of such problems. However, existing work is either designed for the physical network or insufficient to present the concrete reason of packet loss. In this paper, we propose to record and analyze the on-site forwarding condition of packets during packet-level tracing. The cloud-scale overlay network presents great challenges to achieve this goal with its high network complexity, multi-tenant nature, and diversity of root causes. To address these challenges, we present VTrace, an automatic diagnostic system for persistent packet loss over the cloud-scale overlay network. Utilizing the "fast path-slow path" structure of virtual forwarding devices (VFDs), e.g., vSwitches, VTrace installs several "coloring, matching and logging" rules in VFDs to selectively track the packets of interest and inspect them in depth. The detailed forwarding situation at each hop is logged and then assembled to perform analysis with an efficient path reconstruction scheme. Experiments are conducted to demonstrate VTrace's low overhead and quick responsiveness. We share experiences of how VTrace efficiently resolves persistent packet loss issues after deploying it in Alibaba Cloud for over 20 months.

The leakage of driving positions or traces poses a serious privacy threat upon the users in the vehicular ad-hoc networks (VANETs). A series of pseudonym changing approaches have been proposed to achieve unlinkability between the users' identities and their driving information. To investigate the effectiveness of the changing strategies on user anonymity, it is important to stand at the side of an adversary to implement posterior linking between different pseudonyms. In this paper, we remove the assumption of motion models commonly used in existing works, and propose a novel pseudonym linking scheme by focusing on the prediction of acceleration and direction angle. Further, we plug several side information (e.g., road structure, traffic signal) into the proposed scheme to improve the linking performance. Finally, based on five representative pseudonym changing strategies, extensive experiments are conducted to evaluate the performance of the proposed linking scheme. The experimental results show that the side-information assisted pseudonym linking scheme achieves success rates of over 74%.

Executing deep neural networks for inference on the server-class or cloud backend based on the data generated at the edge of the Internet of Things is desirable due primarily to the limited compute power of the edge devices and the need to protect the confidentiality of the inference neural networks. However, such a remote inference scheme incurs concerns regarding the privacy of the inference data transmitted by the edge devices to the curious backend. This article presents a lightweight and unobtrusive approach to obfuscate the inference data at the edge devices. It is lightweight in that the edge device only needs to execute a small-scale neural network; it is unobtrusive in that the edge device does not need to indicate whether obfuscation is applied. Extensive evaluation by three case studies of free-spoken digit recognition, handwritten digit recognition, and American sign language recognition shows that our approach effectively protects the confidentiality of the raw forms of the inference data while effectively preserving backend's inference accuracy.

Recent studies have exploited moving target defense (MTD) for thwarting false data injection (FDI) attacks against the state estimation (SE) by actively perturbing branch parameters (i.e., impedance or admittance) in power grids. To hide the activation of MTD from attackers, a new strategy named hidden MTD has been proposed by the latest literature. A hidden MTD can increase the defender’s chance to detect FDI attacks and avoid the attacker from inferring new branch parameters. However, by using an MTD-confirming detector like the bad data detection (BDD) checker in SE, we observe that it is still possible for the attacker to detect this hidden MTD when the power flows change with time. To uncover the insight of MTD’s hiddenness, we study the conditions needed for achieving a hidable MTD. We find that the hiddenness of MTD is closely related to the branch perturbations, system topology, and attacker’s knowledge. From the attacker’s perspective, we prove that an MTD can be detected by the attacker only if he/she knows the previous parameters of a set of branches that forms a circle and the measurements corresponding to those branches after MTD. But once the attacker has full knowledge of branch parameters before MTD and has obtained all measurements after MTD, it is proved that we can never achieve a hidable and effective MTD. From the defender’s perspective, since it is impossible to know the attacker’s capability, we cannot determine whether a constructed MTD is hidable or not by purely depending on the MTD design. To address this issue, we propose that, by protecting a basic set of measurements, we always can achieve a hidable and effective MTD regardless of the changes of power flows, the attacker’s knowledge, and the branch perturbations. Furthermore, we validate our findings with the IEEE standard test power systems.

Database-driven cognitive radio has been well recognized as an efficient way to reduce interference between Primary Users (PUs) and Secondary Users (SUs). In database-driven cognitive radio, PUs and SUs must provide their locations to enable dynamic channel allocation, which raises location privacy breach concern. Previous studies only focus on unilateral privacy preservation, i.e., only PUs' or SUs' privacy is preserved. In this paper, we propose to protect bilateral location privacy of PUs and SUs. The main challenge lies in how to coordinate PUs and SUs to maximize their utilities provided that their location privacy is protected. We first introduce a quantitative method to calculate both PUs' and SUs' location privacy, and then design a novel privacy preserving Utility Maximization protocol (UMax). UMax allows for both PUs and SUs to adjust their privacy preserving levels and optimize transmit power iteratively to achieve the maximum utilities. Through extensive evaluations, we demonstrate that our proposed protocol can efficiently increase the utilities of both PUs and SUs while preserving their location privacy.

Recent studies have considered thwarting false data injection (FDI) attacks against state estimation in power grids by proactively perturbing branch susceptances. This approach is known as moving target defense (MTD). However, despite of the deployment of MTD, it is still possible for the attacker to launch stealthy FDI attacks generated with former branch susceptances. In this paper, we prove that, an MTD has the capability to thwart all FDI attacks constructed with former branch susceptances only if (i) the number of branches l in the power system is not less than twice that of the system states n (i.e., l ≥ 2n, where n + 1 is the number of buses); (ii) the susceptances of more than n branches, which cover all buses, are perturbed. Moreover, we prove that the state variable of a bus that is only connected by a single branch (no matter it is perturbed or not) can always be modified by the attacker. Nevertheless, in order to reduce the attack opportunities of potential attackers, we first exploit the impact of the susceptance perturbation magnitude on the dimension of the stealthy attack space, in which the attack vector is constructed with former branch susceptances. Then, we propose that, by perturbing an appropriate set of branches, we can minimize the dimension of the stealthy attack space and maximize the number of covered buses. Besides, we consider the increasing operation cost caused by the activation of MTD. Finally, we conduct extensive simulations to illustrate our findings with IEEE standard test power systems.

Advanced traveler information systems (ATISs) provide travelers with public transportation information to improve the quality of individual life and alleviate congestion as well as air pollution. However, existing works have not fully incorporated bike sharing systems within ATIS, providing no interaction with other modalities nor taking bike stocks into account. In addition, the uncertainty of traffic conditions and multimodal routing makes it challenging to accurately estimate the travel time. In this paper, we leverage large-scale historical data collected in London and construct a multimodal transportation network, including bus, tube, public bikes, and walking. We solve the modalities aggregation problem by practically modeling the travel time, arrival time, bike stock, and transfer time between different transport modalities. Furthermore, we propose TERP, a time-event-dependent route planner that optimizes both trip duration and reliability. We conduct experiments on extensive real-world data with over 23 million arrival records and 15 million stock records on more than 10 000 stations from transport for London platform (TfL). The results validates 14.91% reduction of actual total trip duration and 56.28% improvement in terms of route reliability in rush hours comparing with TfL.

Privacy-preserving data aggregation (DA) in network systems, e.g., Internet of Things (IoT), is a challenging problem, considering the dynamic network topology, limited computing capacity, energy supply of IoT devices, etc. The difficulty is exaggerated when there exist dishonest nodes, and how to ensure privacy, accuracy, and robustness of the DA process against dishonest nodes remains an open issue. Different from the widely investigated cryptographic approaches, in this paper, we address this challenging problem by exploiting the distributed consensus technique. To mitigate the pollution from dishonest nodes, we propose an enhanced secure consensus-based DA (E-SCDA) algorithm that allows neighbors to detect dishonest nodes, and derive the error bound when there are undetectable dishonest nodes. We prove the convergence of the E-SCDA and show that the algorithm can preserve the privacy associated to nodes' initial states. Extensive simulations have shown that the proposed algorithm has a high convergence accuracy and low complexity, even when there exist dishonest nodes in the network.

In an industrial cyber-physical system (iCPS), the controller plays a critical role in guaranteeing reliability and stability. Therefore, redundant controller architecture is a well-adopted approach by distributed control systems (DCS), supervisory control and data acquisition (SCADA), and other typical iCPSs. They monitor and control the critical industrial process, such as power generation, chemical industry, water treatment plant, etc. Redundant controller architecture has been designed and largely implemented in response to unpredictable mechanical failures. However, this structure initially proposed for guaranteeing reliability and safety may expand the cyber-attack surface, posing the risk that an attacker may take advantage of this architecture for stealthy attacks. In this article, we analyze the vulnerability arising from the redundant controller architecture and propose a combined attack methodology against these redundant controller architecture systems in a stealthy manner. We find several 0-day vulnerabilities of the real-world devices from three manufacturers and further implement the combined attack over these devices. Our experimental results over various types of real-world devices show that the redundant controller architecture can be exploited to compromise all tested systems stealthily. We also present guidelines for mitigating this risk.

Many clock synchronization protocols based on message passing, e.g., the Network Time Protocol (NTP), assume symmetric network delays to estimate the one-way packet transmission time as half of the round-trip time. As a result, asymmetric network delays caused by either network congestion or malicious packet delays can cause significant synchronization errors. This article exploits sinusoidal voltage signals of an alternating current (AC) power grid to limit the impact of the asymmetric network delays on these clock synchronization protocols. Our extensive measurements show that the voltage signals at geographically distributed locations in a city are highly synchronized. Leveraging calibrated voltage phases, we develop a new clock synchronization protocol that we call Grid Time Protocol (GTP), which allows direct measurement of one-way packet transmission times between its slave and master nodes, subject to an analytic condition that can be easily verified in practice. The direct measurements render GTP resilient against asymmetric network delays under this condition. A prototype implementation of GTP maintains sub-millisecond synchronization accuracy for two nodes tens of kilometers apart in the presence of malicious packet delays. The result has been demonstrated for both Singapore and Hangzhou, China. Simulations driven by real network delay measurements between Singapore and Hangzhou under both normal and congested network conditions also show the synchronization accuracy improvement by GTP. We believe that GTP is suitable for grid-connected distributed systems that are currently served by NTP but desire higher resilience against unfavorable network dynamics and packet delay attacks.

As an innovative mobility strategy, public bike-sharing has grown dramatically worldwide. Though it provides convenient, low-cost, and environmental-friendly transportation, the unique features of bike-sharing systems give rise to problems for both users and operators. The primary issue is the uneven distribution of bikes caused by ever-changing usage and (available) supply. This imbalance necessitates efficient bike rebalancing strategies, which depends highly on bike mobility modeling and prediction. In this paper, a trace-driven simulation-based prediction approach is proposed by simultaneously taking user mobility demand and real-time status of stations into consideration. We extensively evaluate the performance of our design with the dataset from one of the world's largest public bike-sharing systems located in Hangzhou, China, which owns more than 2800 stations. The evaluation results show an 85 percentile relative error of 0.6 for checkout and 0.4 for checkin prediction. The preliminary results on how the predictions can be used for bike rebalancing are also provided. We believe that this new mobility modeling and prediction approach can improve the bike-sharing system operation algorithm design and pave the way for rapid deployment and adoption of bike-sharing systems across the globe.

The rapid development of bike-sharing systems has brought people enormous convenience during the past decade. On the other hand, high transport flexibility gives rise to problems for both users and operators. For users, dynamic distribution of shared bikes caused by uneven user demand often leads to the check in or check out service unavailable at some stations. For operators, unbalanced bike usage comes with more bike broken and growing maintenance cost. In this paper, we consider enhancing user experiences and rebalance bicycle utilization by directing users to different stations with a higher success rate of rental and return. For the first time, we devise a trip advisor that recommends bike check-in and check-out stations with joint consideration of service quality and bicycle utilization. To ensure service quality, we firstly predict the user demand of each station to obtain the success rate of rental and return in the future. Experiments indicate that the precision of our method is as much as 0.826, which has raised by 25.9 percent as compared with that of the historical average method. To rebalance bike usage, from historical data, we identify that biased bike usage is rooted from circumscribed bicycle circulation among few active stations. Therefore, with defined station activeness, we optimize the bike circulation by leading users to shift bikes between highly active stations and inactive ones. We extensively evaluate the performance of our design through real-world datasets. Evaluation results show that the percentage of frequently used bikes decreases by 33.6 percent on usage number and 28.6 percent on usage time.

Broadcast of critical information such as emergency traffic messages in vehicular ad hoc networks (VANETs) has to address jamming with dynamic network topology. In this paper, we propose a deep reinforcement learning based cooperative power control scheme for VANET broadcast against reactive jammers who can observe the ongoing broadcast states. The neural episodic control based cooperative power control scheme uses the convolutional neural network and differentiate neural dictionary to accelerate the learning speed for the VANETs with dynamic topology. Simulation results have shown that the proposed scheme can effectively improve the packet delivery rate and reduce the energy consumption of the broadcast compared with other power control schemes.

The densification deployment of small cells emerging into 5G cellular networks can achieve high capacity, but is faced with the challenge of how to manage energy consumption and inter-cell interference well in time-varying channels. In this paper, we propose a reinforcement learning based downlink power control algorithm to manage interference for the ultra-dense small cell networks. More specifically, base stations of the small cells use Q-learning to select the downlink transmit powers. A transfer learning method called hotbooting is applied to further accelerate the learning speed and save the energy consumption based on the estimated user density without being aware of the network and channel model of the other small cells. Simulation results demonstrate this scheme significantly improves the network throughput and saves the energy consumption compared with the benchmark, a data-driven based transmission power adaptation scheme.

Rapid development of bike-sharing systems has brought people enormous convenience during the past decade. On the other hand, high transport flexibility comes with dynamic distribution of shared bikes, leading to an unbalanced bike usage and growing maintenance cost. In this paper, we consider to rebalance bicycle utilization by means of directing users to different stations. For the first time, we devise a trip advisor that recommends bike check-in and check-out stations with joint consideration of service quality and bicycle utilization. From historical data, we firstly identify that biased bike usage is rooted from circumscribed bicycle circulation among few active stations. Therefore, with defined station activeness, we optimize the bike circulation by leading users to shift bikes between highly active stations and inactive ones. We extensively evaluate the performance of our design through real-world datasets. Evaluation results show that the percentage of frequent used bikes decreases by 33.6% on usage number and 28.6% on usage time.

Increasing dynamics of electrical loads presents uncertainty and hence new challenges for power grid controls and optimization. In economic dispatch control (EDC) for minimizing generation cost, demand reporting by customers is a promising approach for managing the uncertainty, but it raises important privacy concerns. Adding random noise to aggregate queries of demand reports can provide differential privacy (DP) for the individual customers. But the noisy query results can adversely impact the EDC's optimality. In this paper, we analyze the privacy cost in demand reporting in terms of how DP-induced noise will increase the total generation cost. Our analysis shows that the noise amounts for different customers are intricately coupled with one another in determining the total cost. In view of the coupling, we apply the principle of Shapley value to attribute fair shares of the total cost to the power grid buses. For efficient sharing of the privacy cost, in a manner scalable to large power systems with many buses, we additionally propose heuristic algorithms to approximate the Shapley value. Trace-driven simulations based on a 5-bus power system model validate our analysis and illustrate the performance of the proposed cost sharing algorithms.

In this demo, we propose I-Navi, an Indoor Navigation system which leverages the gradient WiFi signal. To be more adaptive to time-variant RSSI and enrich information dimension, I-Navi exploits a three-step backward gradient binary method. Meanwhile, we adopt a lightweight online dynamic time warping (DTW) algorithm to achieve real-time navigation. We fully implemented I-Navi on smartphones and conducted extensive experiments in a five-story campus building and a newly opened two-floor shopping mall with a 90% accuracy of 2m and 3.2m achieved at two places.

Jamming is a typical attack by exploiting the nature of wireless communication. Lots of researchers are working on improving energy-efficiency of jamming attack from the attacker’s view. Whereas, in the low-duty-cycle wireless sensor networks where nodes stay asleep most of time, the design of jamming attack becomes even more challenging especially when considering the stochastic transmission pattern arising from both the clock drift and other uncertainties. In this paper, we propose LearJam, a novel learning-based jamming attack strategy against low-duty-cycle networks, which features the two-phase design consisting of the learning phase and attacking phase. Then in order to degrade the network throughput to the maximal degree, LearJam jointly optimizes these two phases subject to the energy constraint. Moreover, such process of optimization is operated iteratively to accommodate the requirement of practical implementation. Conversely, we also discuss how the state-of-the-art mechanisms can defend against LearJam, which will aid the researchers to improve the security of low-duty-cycle networks. Extensive simulations show that our design achieves significantly higher number of successful attacks and reduces the network’s throughput considerably, especially in a sparse low-duty-cycle network, compared with some typical jamming strategies.

For disconnected Vehicular Ad hoc NETworks (VANETs), the carry-and-forward mechanism is promising to ensure the delivery success ratio at the cost of a longer delay, as the vehicle travel speed is much lower than the wireless signal propagation speed. Estimating delay is critical to select the paths with low delay, and is also challenging given the random topology and high mobility, and the difficulty to let the message propagate along the selected path. In this paper, we first propose a simple yet effective propagation strategy considering bidirectional vehicle traffic for two-dimensional VANETs, so the opposite-direction vehicles can be used to accelerate the message propagation and the message can largely follow the selected path. Focusing on the propagation delay, an analytical framework is developed to quantify the expected path delay. Using the analytical model, a source node can apply the shortest-path algorithm to select the path with the lowest expected delay. Performance evaluation by simulation show that, when the vehicle density is uneven but known, the proposed Minimum Delay Routing Algorithm can achieve a substantial reduction in delay compared with the geocast-routing approach, and its performance is close to the flooding-based Epidemic algorithm, while our solution maintains only a single copy of the message.

Electric bikes (e-bikes) are booming in China, providing a flexible and energy-efficient trip mode for people all around the country. However, the problems such as traffic accidents and disrupting social order caused by e-bikes bring much worry to the government operators. And the behavior of e-bike users also attracts the sociological researchers' interest due to their characteristics (e.g. lower income and higher mobility). In order to facilitate government operators in monitoring e-bikes and to help sociological researchers study the behavior of e-bike users, we develop a data analysis and visualization system based on large-scale e-bike data. We analyze the e-bike mobility and e-bike user behavior, and design a visual interface that allows interaction with the analysis results. Our system not only provides the function of e-bike monitoring but also serves as a platform for mobility analysis and user behavior analysis.

As an innovative mobility strategy, public bike-sharing has grown dramatically worldwide. Though providing convenient, low-cost and environmental-friendly transportation, the unique features of bike-sharing systems give rise to problems to both users and operators. The primary issue among these problems is the uneven distribution of bicycles caused by the ever-changing usage and (available) supply. This bicycle imbalance issue necessitates efficient bike re-balancing strategies, which depends highly on bicycle mobility modeling and prediction. In this paper, for the first time, we propose a spatio-temporal bicycle mobility model based on historical bike-sharing data, and devise a traffic prediction mechanism on a per-station basis with sub-hour granularity. We extensively evaluated the performance of our design through a one-year dataset from the world's largest public bike-sharing system (BSS) with more than 2800 stations and over 103 million check in/out records. Evaluation results show an 85 percentile relative error of 0.6 for both check in and check out prediction. We believe this new mobility modeling and prediction approach can advance the bike re-balancing algorithm design and pave the way for the rapid deployment and adoption of bike-sharing systems across the globe.

Wireless rechargeable sensor networks have recently emerged as a promising platform that can effectively solve the power constraint problem suffered by traditional battery powered systems. The problem of determining the best charging routes for maximizing charging efficiency has been studied extensively. However, the task assignment problem, which plays a crucial role in efficiently utilizing the harvested energy and thus minimize the charging delay, has received rather limited attention. In this paper, we study the problem of assigning a given set of tasks in a wireless rechargeable sensor network while maximizing the charger's velocity to minimize the charging delay. We first propose an online task assignment algorithm, namely Lower Bound assignment (LB), that yields a quantifiable lower bound on the charging velocity while guaranteeing a feasible assignment. This algorithm further enables the transformation of our considered task assignment problem into a variation of the classical multiple knapsack problem. We then present a fully polynomial-time approximation scheme with a (2+ε)-approximation ratio, namely ACT, that is built upon an existing greedy algorithm designed for the original knapsack problem. Extensive experimental results presented herein demonstrate that ACT is able to achieve near-optimal performance in most cases, and can achieve more than 15% performance improvement compared to the baseline algorithms.

Most existing works investigate consensus with noise following a certain distribution, e.g., Gaussian distribution, with fixed expectation and variance, which may not be satisfied in practical applications. This paper investigates the discrete system consensus under bounded noise, which is important and practical problem. We first provide necessary and sufficient conditions for the convergence of consensus under bounded noise. To be more general, we derive an analytical bound to show the max-min difference between the nodes' states when the general consensus algorithm converges to a stable state. Then, a novel consensus algorithm, fast consensus under bounded noise (FCBN), is proposed to eliminate the accumulative error caused by the bounded noise. It is proved that FCBN has a faster convergence speed and a higher consensus accuracy than general consensus algorithms. Extensive simulations demonstrate the effectiveness of the proposed algorithm.

Minimizing the end-to-end delay for data dissemination in a large-scale VANET with both buses of fixed schedules and taxis of random schedules is a challenging issue, due to the scalability, high-mobility, and network heterogeneity concerns. Particularly, the mix of random taxis and fixed-scheduled buses makes the delay components along a path dependent and hard to estimate. In this paper, to address the scalability and high-mobility issues, we introduce a store-and-forward framework for VANETs with extra storage using “drop boxes”, which function similar to network routers. Next, we propose an optimal link strategy which is independent of the message arrival time and can be executed in a distributed manner. Then, we derive the expected path delay, considering the dependence of the delay components along the path, and propose the optimal routing strategy to minimize the expected path delay. Trace-driven simulations have been used to validate the rigorous analysis, and demonstrate the superior performance of the proposed strategies, which result in a substantial delay reduction and a much higher delivery ratio when compared with the state-of-the-art solutions without drop boxes. The strategies can further improve the delay performance when compared with the over-simplified routing solutions which ignore the dependence of the delay components.

Limited energy in each node is the major design constraint in wireless sensor networks (WSNs). To overcome this limit, wireless rechargeable sensor networks (WRSNs) have been proposed and studied extensively over the last few years. In a typical WRSN, batteries in sensor nodes can be replenished by a mobile charger that periodically travels along a certain trajectory in the sensing area. To maximize the charged energy in sensor nodes, one fundamental question is how to control the traveling velocity of the charger. In this paper, we first identify the optimal velocity control as a key design objective of mobile wireless charging in WRSNs. We then formulate the optimal charger velocity control problem on arbitrarily-shaped irregular trajectories in a 2D space. The problem is proved to be NP-hard, and hence a heuristic solution with a provable upper bound is developed using novel spatial and temporal discretization. We also derive the optimal velocity control for moving the charger along a linear (1D) trajectory commonly seen in many WSN applications. Extensive simulations show that the network lifetime can be extended by 2.5× with the proposed velocity control mechanisms.

Recently, the industrial wireless protocols have been widely used around the world. However, the unreliable communication media between the sensors and the central controller renders the wireless signal channel vulnerable to many attacks. Various efforts have been devoted to study the influence of specific malicious attacks from the aspect of theoretical investigation based on different assumptions. This paper focuses on verifying the optimal Denial-of-Service (DoS) jamming attack strategy on a class of wireless industrial control system from the view of experiments. We first introduce typical control system model and DoS attack model, and an optimal DoS attack schedule against LQG control based on these models. Then, we establish a semi-physical security testbed which consists of virtual plant, physical controller and communication process. We also realize wireless DoS attacks by exploiting the USRP device. Through extensive experiments and analysis, we investigate the performance of different DoS attack strategies on the LQG control system over an inverted pendulum.

Database-driven cognitive radio has been well recognized as an efficient way to reduce interference between Primary Users (PUs) and Secondary Users (SUs). In database-driven cognitive radio, PUs and SUs must provide their locations to enable dynamic channel allocation, which raises location privacy breach concern. Previous studies only focus on unilateral privacy preservation, i.e., Only PUs' or SUs' privacy is preserved. In this paper, we propose to protect bilateral location privacy of a PU and an SU. The main challenge lies in how to coordinate the PU and SU to maximize their utility provided that their location privacy is protected. We first introduce a quantitative method to calculate both PU's and SU's location privacy, and then design a novel privacy preserving Utility Maximization protocol (UMax). UMax allows for both PU and SU to adjust their privacy preserving levels and optimize transmit power iteratively to achieve the maximum utility. Through extensive evaluations, we demonstrate that our proposed mechanism can efficiently increase the utility of both PU and SU while preserving their location privacy.

In this demo, we propose Phone meter which leverages the RF energy harvesting technologies to measure the strength of Electromagnetic Field (EMF). To this end, Phone meter combines EMF sensor with the smartphone through audio interface without any modifications to the phone. We fully implement the low-cost Phone meter and conduct extensive experiments to prove the functionality of Phone meter. Phone meter achieves about 13:7% relative error in average compared with the industrial-grade spectrum analyzer with significantly reduced the costs.

Considering an optimal investment problem for a retailer in electricity market, the objective is to seek the optimal investment decision that maximizes the weighted sum of the expected return and the variance of wealth. Unlike existing works, the price fluctuation of both the wholesale and retail side of electricity market is considered, and the retailer can invest its wealth in electricity market and traditional financial market simultaneously. Hence, there is a complicated wealth dynamic, which is the main challenge in our work. In this paper, by utilizing the method of Lagrange multiplier and the classical Tchebycheff inequality, we first show that the investment problem is a quadratic programming problem in terms of the decision variable, and thus has a unique optimal solution. Then, a closed-form optimal solution is derived by solving the stationary equation and comparing the feasible solution interval. Based on the optimal solution, we find the key price, which will affect the investment is the wholesale price rather than the retail price. Moreover, with a similar analysis approach, we also provide the optimal solution considering a more general model, which allows the retailer to purchase the electricity temporarily to avoid the supply shortage. Extensive simulations demonstrate the better performance of the proposed solution over the Kelly strategy widely used in the financial market.

The wireless rechargeable sensor network is a promising platform for long-term applications such as inventory management, supply chain monitoring, and so on. For these applications, sensor localization is one of the most fundamental challenges. Different from a traditional sensor node, a wireless rechargeable sensor has to be charged above a voltage level by the wireless charger in order to support its sensing, computation, and communication operations. In this work, we consider the scenario where a mobile charger stops at different positions to charge sensors and propose a novel localization design that utilizes the unique Time of Charge (TOC) sequences among wireless rechargeable sensors. Specifically, we introduce two efficient region dividing methods, Internode Division and Interarea Division , to exploit TOC differences from both temporal and spatial dimensions to localize individual sensor nodes. To further optimize the system performance, we introduce both an optimal charger stop planning algorithm for the single-sensor case and a suboptimal charger stop planning algorithm for the generic multisensor scenario with a provable performance bound. We have extensively evaluated our design by both testbed experiments and large-scale simulations. The experiment and simulation results show that by as less as five stops, our design can achieve sub-meter accuracy and the performance is robust under various system conditions.

Abstract This paper considers the problem of localizing a group of targets whose number is unknown by wireless sensor networks. At each time slot, to save energy and bandwidth resources, only part of sensor nodes are scheduled to activate to remain continuous monitoring of all the targets. The localization problem is formulated as a sparse vector recovery problem by utilizing the spatial sparsity of targets’ location. Specifically, each activated sensor records the RSS values of the signals received from the targets and sends the measurements to the sink node where a compressive sampling‐based localization algorithm is conducted to recover the number and locations of targets. We decompose the problem into two sub‐problems, namely, which sensor nodes to activate, and how to utilize the measurements. For the first subproblem, to reduce the effect of measurement noise, we propose an iterative activation algorithm to re‐assign the activation probability of each sensor by exploiting the previous estimate. For the second subproblem, to further improve the localization accuracy, a sequential recovery algorithm is proposed, which conducts compressive sampling on the least squares residual of the previous estimate such that all the previous estimate can be utilized. Under some mild assumptions, we provide the analytical performance bound of our algorithm, and the running time of proposed algorithm is given subsequently. Simulation results demonstrate the effectiveness of our algorithms.Copyright © 2013 John Wiley & Sons, Ltd.

In wireless rechargeable sensor networks (WRSN), rechargeable sensor nodes harvest ambient RF energy from power sources such as the RFID readers. However, the simultaneous transmissions may cause severe communication collisions. Different from traditional approaches which mainly resolve such collisions at the MAC layer, in this work we optimize the deployment of RFID reader in order to avoid the communication collisions by exploiting the differences in the time of charge among rechargeable sensor nodes. Specifically, when the reader is able to cover the whole deployment field, an efficient collision-free solution with proved optimality is presented to minimize the communication delay in the network. Simulation results are employed to verify the proposed algorithm.

Recently, much research effort has been devoted to employing mobile chargers for energy replenishment of the robots in robotic sensor networks. Observing the discrepancy between the charging latency of robots and charger travel distance, we propose a novel tree-based charging schedule for the charger, which minimizes its travel distance without causing the robot energy depletion. We analytically evaluate its performance and show its closeness to the optimal solutions. Furthermore, through a queue-based approach, we provide theoretical guidance on the setting of the remaining energy threshold at which the robots request energy replenishment. This guided setting guarantees the feasibility of the tree-based schedule to return a depletion-free charging schedule. The performance of the tree-based charging schedule is evaluated through extensive simulations. The results show that the charger travel distance can be reduced by around 20%, when compared with the schedule that only considers the robot charging latency.

Wireless rechargeable sensor network is a promising platform for long-term applications such as inventory management, supply chain monitoring and so on. For these applications, sensor localization is one of the most fundamental challenges. Different from traditional sensor node, wireless rechargeable sensor has to be charged above a voltage level by the wireless charger in order to support its sensing, computation and communication operations. In this work, we consider the scenario where a mobile charger stops at different positions to charge sensors, and propose a novel localization design that utilizes the unique Time of Charge (TOC) sequences among wireless rechargeable sensors. Specifically, we introduce two efficient region dividing methods, Inter-node Division and Inter-area Division, to exploit TOC differences from both temporal and spatial dimensions to localize individual sensor nodes. To further optimize the system performance, we introduce both an optimal charger stop planning algorithm for single sensor case and a suboptimal charger stop planning algorithm for the generic multisensor scenario with a provable performance bound. We have extensively evaluated our design by both testbed experiments and large-scale simulations. The experiment and simulation results show that by as less as 5 stops, our design can achieve sub-meter accuracy and the performance is robust under various system conditions.

Low-duty-cycle network plays an crucial role in improving energy efficiency of wireless communication, where nodes stay asleep most of time. Despite energy saving, the security of low-duty-cycle networks is of great concern. The attacking strategy design becomes even more challenging considering the stochastic transmission patterns arising from both the clock drift and other uncertainties. In this paper, we propose LearJam, a novel two-phase energy-efficient learning-based jamming attack strategy against low-duty-cycle networks, where the attacker estimates the distribution of transmission period in the learning phase, and schedules its jamming attacks in the attacking phase based on this estimated distribution. We jointly optimize the learning duration and the attacking duration under the energy constraint in order to degrade the network throughput to the maximal degree. We propose simple yet effective methods to solve both the single-node and multi-node scenarios. We further discuss a state-of-the-art mechanism defending against LearJam by re-scheduling transmission pattern, which will aid the researchers to improve the security of low-duty-cycle networks. Extensive simulations show that our design achieves significantly higher number of successful attacks (increasing 38%-762%) in a sparse low-duty-cycle network compared with some traditional jamming strategies.

Different from energy harvesting which generates dynamic energy supplies, the mobile charger is able to provide stable and reliable energy supply for sensor nodes, and thus enables sustainable system operations. While previous mobile charging protocols either focus on the charger travel distance or the charging delay of sensor nodes, in this work we propose a novel Energy Synchronized Charging (ESync) protocol, which simultaneously reduces both of them. Observing the limitation of the Traveling Salesman Problem (TSP)-based solutions when nodes energy consumptions are diverse, we construct a set of nested TSP tours based on their energy consumption rates, and only nodes with low remaining energy are involved in each charging round. Furthermore, we propose the concept of energy synchronization to synchronize the charging requests sequence of nodes with their sequence on the TSP tours.

Different from energy harvesting which generates dynamic energy supplies, the mobile charger is able to provide stable and reliable energy supply for sensor nodes, and thus enables sustainable system operations. While previous mobile charging protocols either focus on the charger travel distance or the charging delay of sensor nodes, in this work we propose a novel Energy Synchronized Charging (ESync) protocol, which simultaneously reduces both of them. Observing the limitation of the Traveling Salesman Problem (TSP)-based solutions when nodes energy consumptions are diverse, we construct a set of nested TSP tours based on their energy consumptions, and only nodes with low remaining energy are involved in each charging round. Furthermore, we propose the concept of energy synchronization to synchronize the charging re- quests sequence of nodes with their sequence on the TSP tours. Experiment and simulation demonstrate ESync can reduce charger travel distance and nodes charging delay by about 30% and 40% respectively.

The Wireless Identification and Sensing Platform (WISP) has become a very promising experimental platform of wireless rechargeable sensor networks (WRSN), which integrates the sensing and computation capabilities to the traditional RFID tags. In such kind of networks, the simultaneous transmission may introduce severe communication collisions, which have attracted various research efforts for resolving such collisions at the MAC layer. However, different from existing works, we avoid such communication collisions through proper reader movement by exploiting the differences in the time of charge among rechargeable sensor nodes. We formulate the optimization problem and prove that complexity of the optimal solution is NP-hard, and propose a simple yet effective algorithm to optimize both the reader stop location and stop time for minimizing the total communication delay. Extensive simulation under different system settings show that our design can largely reduce the communication delay and outperform the baseline design by at least 20%.

Smartphones and mobile devices have greatly penetrated the daily lives of many people. While participatory/pervasive sensing has gained wide adoptions by leveraging various onboard sensors on mobile devices, another powerful resource, the computational power on these mobile devices has been less frequently harnessed by researchers and practitioners. To fill this gap, we propose in this work the modeling, analysis, and implementation of participatory computing. Specifically, we propose REPC, a generic randomized task assignment framework for the participatory computing paradigm, which guarantees the overall system performance with close to minimal workload at individual participating devices. To achieve these design objectives, we model the intrinsic relationship between the workload of individual devices and the probability they complete their assigned tasks. Based on our modeling results, we analyze the maximal system capacity for any given participatory computing system and derive the minimal workload for individual participating devices to achieve the overall system performance requirement. We have fully implemented our design on the Android platform and demonstrated its performance through a representative participatory computing application. Extensive experiments and simulation results demonstrate that our design is able to achieve more than 90% task completion ratios with only 10% system overhead in practice.

Integrated with low-power micro-controllers and sensors, RFID-based wireless rechargeable sensor node is a very promising platform for applications such as inventory management, supply chain monitoring etc. Among other major research challenges, one of the most essential problems in such wireless rechargeable sensor networks is how to minimize the communication delay among RFID readers and RFID-based rechargeable nodes. While the existing works have mostly focused on the collision avoidance among RFID-based nodes, in this work we study an orthogonal approach which focuses on how to optimally plan the movement of the reader so as to minimize the communication delay in the network. To solve this problem, we introduce both an optimal solution for the linear reader movement pattern and an approximation solution for the generic two-dimensional reader move pattern with a provable approximation ratio. In addition, we also provide a solution for guaranteeing the quality of communication while minimizing the communication delay. We verify our observations through testbed experiments and extensively evaluate our design by both emulations and large-scale simulations. The results show our design can effectively reduce communication delay in wireless rechargeable sensor networks when compared with baseline solutions.

We investigate the state estimation problem in cyber-physical systems (CPS) where the dynamical physical process is measured by a wireless sensor and the measurements are transmitted to a remote state estimator. It has been shown that the estimation performance strongly depends on the wireless communication quality. To enhance the estimation performance, we apply the cognitive radio technique to the system and propose a CHAnnel seNsing and switChing mEchanism (CHANCE) to explore opportunistic accessibility of multiple channels. We consider two types of wireless channels, i.e., one unlicensed channel which can be accessed freely and several licensed channels which have been pre-assigned to primary users. For the single-licensed-channel case, we develop a necessary condition for the estimation stability based on the physical process dynamics, channel quality and the channel sensing accuracy. This condition becomes also sufficient under certain conditions. We also derive the conditions under which the estimation performance is guaranteed to be improved by CHANCE. The above results are then extended to multi-licensed-channel cases. Simulations based on a particular linear system show that, the long-run mean estimation error covariance with CHANCE is at least 63% less than that without CHANCE. It is also shown that CHANCE outperforms the existing RANDOM mechanism in terms of estimation performance.

Time synchronization is a fundamental requirement for the wide spectrum of applications with wireless sensor networks (WSNs). However, most existing time synchronization protocols are likely to deteriorate or even to be destroyed when the WSNs are attacked by malicious intruders. This paper is concerned with secure time synchronization for WSNs under message manipulation attacks. Specifically, the theoretical analysis and simulation results are first provided to demonstrate that the maximum consensus based time synchronization (MTS) protocol would be invalid under message manipulation attacks. Then, a novel secured maximum consensus based time synchronization (SMTS) protocol is proposed to detect and invalidate message manipulation attacks. Furthermore, we prove that SMTS is guaranteed to converge with simultaneous compensation of both clock skew and offset. Extensive numerical results show the effectiveness of our proposed protocol.

We consider the problem of event capture by a rechargeable sensor network. We assume that the events of interest follow a renewal process whose event inter-arrival times are drawn from a general probability distribution, and that a stochastic recharge process is used to provide energy for the sensors' operation. Dynamics of the event and recharge processes make the optimal sensor activation problem highly challenging. In this paper we first consider the single-sensor problem. Using dynamic control theory, we consider a full-information model in which, independent of its activation schedule, the sensor will know whether an event has occurred in the last time slot or not. In this case, a simple and optimal greedy policy for the solution is developed. We then further consider a partial-information model where the sensor knows about the occurrence of an event only when it is active. This problem falls into the class of partially observable Markov decision processes (POMDP). Since the POMDP's optimal policy has exponential computational complexity and is intrinsically hard to solve, we propose an efficient heuristic clustering policy and evaluate its performance. Finally, our solutions are extended to handle a network setting in which multiple sensors collaborate to capture the events. We also provide extensive simulation results to evaluate the performance of our solutions.

Recently, various consensus-based protocols have been developed for time synchronization in wireless sensor networks. However, due to the uncertainties lying in both the hardware fabrication and network communication process, it is not clear how most of the protocols will perform in real implementations. In order to reduce such gap, this paper investigates whether and how the typical consensus-based time synchronization protocols can tolerate the uncertainties in practical sensor networks through extensive testbed experiments. For two typical protocols, i.e., Average Time Synchronization (ATS) and Maximum Time Synchronization (MTS), we first analyze how the time synchronization accuracy will be affected by various uncertainties in the system. Then, we implement both protocols on our sensor network testbed consisted of Micaz nodes. We further investigate the time synchronization performance and robustness under various settings. The extensive experimental results demonstrate the advantages of MTS over ATS.

As a pioneering experimental platform of wireless rechargeable sensor networks, the Wireless Identification and Sensing Platform (WISP) is an open-source platform that integrates sensing and computation capabilities to the traditional RFID tags. Different from traditional tags, a RFID-based wireless rechargeable sensor node needs to charge its onboard energy storage above a threshold in order to power its sensing, computation and communication components. Consequently, such charging delay imposes a unique design challenge for deploying wireless rechargeable sensor networks. In this paper, we tackle this problem by planning the optimal movement strategy of the RFID reader, such that the time to charge all nodes in the network above their energy threshold is minimized. We first propose an optimal solution using the linear programming method. To further reduce the computational complexity, we then introduce a heuristic solution with a provable approximation ratio of (1 + θ)/(1 - ε) by discretizing the charging power on a two-dimensional space. Through extensive evaluations, we demonstrate that our design outperforms the set-cover-based design by an average of 24.7% while the computational complexity is O((N/ε)2).

Participatory Computing is a promising solution to fully utilize the wasted computation resources of mobile devices such as smartphones. In this demo abstract, we present our design and implementation of an participatory computing enhanced pervasive surveillance system. Our evaluation results show that the proposed system can effectively utilize the computation capability of mobile devices while guaranteeing the service reliability even with the intermittent nature of participatory computing.

In this paper, we address the problem of peer-to-peer networking for data dissemination among actors in wireless sensor and actor networks (WSANs), which consist of static sensors, responsible for environment monitoring, and mobile actors, in charge of data collection and task performing. This problem has not been received much attention although peer-to-peer networking has achieved great successes in other networks such as the Internet and mobile ad hoc networks (MANETs). Unlike the Internet and MANETs, WSANs contain static sensors that are energy-constrained and actors that cannot communicate with each other directly. These unique characteristics make the data dissemination problem in WSANs extremely challenging. We present an Energy-Efficient Message Dissemination protocol (EMD) to solve this problem in delay-tolerant WSANs. EMD is grounded on a novel principle of "Carry-Disseminate-Store-and-Forward" proposed for the first time here. While traveling, a source actor disseminates messages (data) to sensors upon contact, which will store the messages and forward them to other actors when they come into communication range. The actors receiving the messages from sensors work as source actors and help to distribute the messages. We theoretically analyze the data dissemination strategy under which the original source actor can distribute its messages to all other actors at minimum communication cost within a given delay bound. Through extensive simulations we demonstrate the performance of EMD.

Networked cyber-physical systems (NCPS), where control and communication are closely integrated, have been envisioned to have a large number of high-impact applications. In this paper, a joint optimization framework is presented, which combines the objective of control as well as other relevant system objectives and constraints such as communication errors, delays and the limited capabilities (e.g., energy capacities) of devices. The problem is solved by an online optimization approach, which consists of a communication protocol and a simulated annealing based control algorithm. Meanwhile, by taking into account the communication cost, we optimize the control intervals by integrating two kinds of acceptances, i.e., cyber and physical acceptances, into the control algorithm. Numerical results show the effectiveness of the proposed approach.

Wireless Rechargeable Sensor Network (WRSN) is an emerging technology to address the energy constraint in sensor networks. The protocol design in WRSN is extremely challenging due to the complicated interactions between rechargeable sensor nodes and readers, capable of mobility and functioning as energy distributors and data collectors. In this paper, we for the first time investigate the optimal scheduling problem in WRSN for stochastic event capture, i.e., how to jointly mobilize the readers for energy distribution and schedule sensor nodes for optimal quality of monitoring (QoM). We analyze the QoM for three application scenarios: i) the reader travels at a fixed speed to recharge sensor nodes and sensor nodes consume the collected energy in an aggressive way, ii) the reader stops to recharge sensor nodes for a predefined time during its periodic traveling and sensor nodes deplete energy aggressively, iii) the reader stops to recharge sensor nodes but sensor nodes can adopt optimal duty cycle scheduling for maximal QoM. We provide analytical results for achieving the optimal QoM under arbitrary parameter settings. Extensive simulation results are offered to demonstrate the correctness and effectiveness of our results.

As technology advancements in robotics and wireless communication, tracking mobile targets using mobile sensors has aroused widespread concern in recent years. In this paper, we propose a novel coordinative moving strategy for autonomous mobile sensor networks to guarantee the target can be detected in each observed step while minimizing the amount of moving sensors. The proposed scheme consists of obtaining the current position of the target, which is then used to predict the next time-step location of the target. Once the uncertainty region of the target's position is defined, the proposed method allows the mobile sensors to cover it in an optimal way. Therefore, we can assign each mobile sensor to an optimal location to cover the uncertainty region while minimizing the total traveled distance of sensors. Extensive simulations are given to evaluated performance and demonstrate the efficiency of the proposed strategy.

In this paper, we study power allocation in cognitive sensor networks where cognitive users (cognitive enabled sensor nodes) opportunistically share a common spectrum with primary users (licensed devices). We define an energy efficiency-oriented utility function as a new metric to evaluate power allocation. Consider that sensor nodes are self-interested to maximize their own utility, we formulate the energy efficient power allocation problem as a non-cooperative game. We firstly prove that there exist Nash equilibriums in the proposed game. Secondly, we prove that the power allocation game is a supermodular game with some conditions. Finally, we use best response algorithm to identify the Nash equilibrium. Simulations are conducted to demonstrate that the proposed power allocation strategy can achieve satisfactory performance in terms of energy efficiency, convergence speed and fairness in cognitive sensor networks.

In wireless sensor and actuator networks (WSAN), both sensory measurements and control signals transmitted through the wireless media are prone to packet losses. Moreover, the distance between the sender and receiver is a critical factor for the loss rate. Therefore, where to place the controller to ensure optimal control performance is an interesting problem. In this paper, we focus on WSAN with one sensor and one actuator residing at different geographic locations. If the controller is constrained at either the sensor side or actuator side, we derive the necessary and sufficient conditions under which the optimal controller location can be directly determined. For the more general case when the controller can be placed anywhere, with mild assumptions on the packet drop model, we also provide the conditions under which the optimal controller location is unique and can be determined. Numerical simulations based on a practical packet loss model verify our results.

We consider the problem of event capture by a rechargeable sensor network. We assume that the events of interest follow a renewal process whose event inter-arrival times are drawn from a general probability distribution, and that a stochastic recharge process is used to provide energy for the sensors' operation. Dynamics of the event and recharge processes make the optimal sensor activation problem highly challenging. In this paper we first consider the single-sensor problem. Using dynamic control theory, we consider a full-information model in which, independent of its activation schedule, the sensor will know whether an event has occurred in the last time slot or not. In this case, the problem is framed as a Markov decision process (MDP), and we develop a simple and optimal policy for the solution. We then further consider a partial-information model where the sensor knows about the occurrence of an event only when it is active. This problem falls into the class of partially observable Markov decision processes (POMDP). Since the POMDP's optimal policy has exponential computational complexity and is intrinsically hard to solve, we propose an efficient heuristic clustering policy and evaluate its performance. Finally, our solutions are extended to handle a network setting in which multiple sensors collaborate to capture the events. We provide extensive simulation results to evaluate the performance of our solutions.

In this paper, we consider using actors to maintain the quality of sensing in the wireless sensor networks. Due to factors such as battery drainage or physical malfunctions, the number of available sensors normally decreases over time after initial deployment, resulting in performance degradation. To maintain the quality of sensing in the network, actors can be used to allocate spare sensors to sensor-deficient regions (sensor allocation) or to relocate sensors from sensor-abundant regions to sensor-deficient regions (sensor relocation). We first focus on the sensor allocation problem. We introduce a baseline centralized greedy algorithm (GA) for sensor allocation, where global sensor information is communicated to obtain the optimal solution. As GA is only efficient for small networks, we proceed to design a distributed patrolling algorithm for achieving global optimization (DPAG) by using only local information. We then extend our work to the application scenario of sensor relocation by proposing a modified GA and DPAG (M-GA and M-DPAG), respectively. Extensive simulation results are provided to demonstrate the performance of the proposed algorithms.

In target tracking applications, active ultrasonic sensors can provide satisfactory distance estimations, but also suffer from inter-sensor-interference when they are not well scheduled. In this paper, we propose a distributed saturation degree based algorithm (DSDA), which assigns the TDMA slot in wireless sensor networks distributively in order to avoid the interference. By adopting a graph coloring technique, Saturation Degree Heuristic, this new algorithm can provide near-optimal slot number in a totally distributed way. Simulation results demonstrate the efficiency of DSDA in terms of slot number, system scalability, tracking accuracy and energy consumption.

Recently, wireless recharging technologies have merged as a promising approach to address the energy constraint problem in Wireless Sensor Networks (WSNs). Far from other energy-harvesting sensor nodes, wireless rechargeable sensor nodes are thin small-size, enabling a large range of applications such as embedded infrastructure sensing and human activity recognition. A typical Wireless Rechargeable Sensor Network (WRSN) includes two components: i) a collection of rechargeable sensor nodes and ii) several readers, capable of mobility and functioning as energy distributors and data collectors. In this paper, we for the first time investigate the optimal scheduling problem in WRSN for stochastic event capture, i.e., how to jointly mobilize the readers for energy distribution and schedule sensor nodes for efficient event capture. We extensively study the problem and analyze the quality of capture for different application scenarios. At last, numerical results are offered to demonstrate the correctness and effectiveness of our solutions.

In this paper, we consider the problem of sensor scheduling with limited resources. Two sensors are used to measure the state of a discrete-time linear process. We assume that each sensor has a maximum duty cycle and at most one sensor can communicate with a remote estimator at each time step due to the limited communication bandwidth. When a sensor is scheduled to send data, it sends the most recent D measurement data to the estimator. Upon receiving the measurement data from the sensors, the estimator computes the optimal estimate of the state of the process. We first present a necessary condition for a sensor scheduling scheme to be optimal. Based on this necessary condition, we construct an optimal scheduling scheme that minimizes the estimation error at the estimator and at the same time satisfies the energy and communication bandwidth constraints. We also provide a sufficient condition on the minimum D such that an optimal scheduling scheme can be constructed. Examples are provided throughout the paper to demonstrate the results developed.