Biblio

This note studies self-triggered tracking control of underactuated surface vessels considering both unknown model dynamics and stochastic noise, where the measured states in the sensors are intermittently transmitted to the controller decided by the triggering condition. While the multi-layer neural network (NN) serves to approximate the unknown model dynamics, a self-triggered adaptive neural model is fabricated to direct the design of control laws. This setup successfully solves the ``jumps of virtual control laws'' problem, which occurs when combining the event-triggered control (ETC) with the backstepping method, seeing [1]–[4]. Moreover, the adaptive model can act as the filter of states, such that the complicated analysis and control design to eliminate the detrimental influence of stochastic noise is no longer needed. Released from the continuous monitoring of the controller, the devised triggering condition is located in the sensors and designed to meet the requirement of stability. All the estimation errors and the tracking errors are proved to be exponentially mean-square (EMS) bounded. Finally, a numerical experiment is conducted to corroborate the proposed strategy.

Nowadays, we have witnessed the early progress on learning the association between voice and face automatically, which brings a new wave of studies to the computer vision community. However, most of the prior arts along this line (a) merely adopt local information to perform modality alignment and (b) ignore the diversity of learning difficulty across different subjects. In this paper, we propose a novel framework to jointly address the above-mentioned issues. Targeting at (a), we propose a two-level modality alignment loss where both global and local information are considered. Compared with the existing methods, we introduce a global loss into the modality alignment process. The global component of the loss is driven by the identity classification. Theoretically, we show that minimizing the loss could maximize the distance between embeddings across different identities while minimizing the distance between embeddings belonging to the same identity, in a global sense (instead of a mini-batch). Targeting at (b), we propose a dynamic reweighting scheme to better explore the hard but valuable identities while filtering out the unlearnable identities. Experiments show that the proposed method outperforms the previous methods in multiple settings, including voice-face matching, verification and retrieval.

Particle filtering is an indispensable method for non-Gaussian state estimation, but it has some problems, such as particle degradation and requiring a large number of particles to ensure accuracy. Biota intelligence algorithms led by Cuckoo (CS) and Firefly (FA) have achieved certain results after introducing particle filtering, respectively. This paper respectively in the two kinds of bionic algorithm convergence factor and adaptive step length and random mobile innovation, seized the cuckoo algorithm (CS) in the construction of the initial value and the firefly algorithm (FA) in the iteration convergence advantages, using the improved after the update mechanism of cuckoo algorithm optimizing the initial population, and will be updated after optimization way of firefly algorithm combined with particle filter. Experimental results show that this method can ensure the diversity of particles and greatly reduce the number of particles needed for prediction while improving the filtering accuracy.

This work proposes an event-triggered adaptive asymptotic tracking control scheme for flexible robotic manipulators. Firstly, by employing the command filtered backstepping technology, the ``explosion of complexity'' problem is overcame. Then, the event-triggered strategy is utilized which makes that the control input is updated aperiodically when the event-trigger occurs. The utilized event-triggered mechanism reduces the transmission frequency of computer and saves computer resources. Moreover, it can be proved that all the variables in the closed-loop system are bounded and the tracking error converges asymptotically to zero. Finally, the simulation studies are included to show the effectiveness of the proposed control scheme.

In most correlation filter target tracking algorithms, poor accuracy in the tracking process for complex field images of the target and scale change problems. To address these issues, this paper proposes an algorithm of adaptive multi-feature fusion with scale change correlation filtering tracking. Our algorithm is based on the rapid and simple Kernel-Correlated Filtering(K CF) tracker, and achieves the complementarity among image features by fusing multiple features of Color Nmae(CN), Histogram of Oriented Gradient(HOG) and Local Binary Pattern(LBP) with weights adjusted by visual evaluation functions. The proposed algorithm introduces scale pooling and bilinear interpolation to adjust the target template size. Experiments on the OTB-2015 dataset of 100 video frames are compared with several trackers, and the precision and success ratio of our algorithm on complex scene tracking problems are 17.7% and 32.1 % respectively compared to the based-KCF.

The output feedback problem of finite-time command filtering for nonlinear systems with random disturbance is addressed in this paper. This is the first time that command filtering and output feedback are integrated so that a nonlinear system with random disturbance converge rapidly in finite time. The uncertain functions and unmeasured states are estimated by the fuzzy logic system (FLS) and nonlinear state observer, respectively. Based on the adaptive framework, command filtering technology is applied to mitigate the problem of ``term explosion'' inherent in traditional methods, and error compensation mechanism is considered to improve the control performance of the system. The developed output feedback controller ensures the boundedness of all signals in the stochastic system within a finite time, and the convergence residual can converge to a small region. The validity of this scheme is well verified in a numerical example.

Due to the insufficient awareness of decision makers on the security risks of industrial cyber-physical systems(ICPS) under cyber-attacks, it is difficult to take effective defensive measures according to the characteristics of different cyber-attacks in advance. To solve the above problem, this paper gives a qualitative analysis method of ICPS security risk from the perspective of defenders. The ICPS being attacked is modeled as a dynamic closed-loop fusion model where the mathematical models of the physical plant and the feedback controller are established. Based on the fusion model, the disruption resources generated by attacks are mathematically described. Based on the designed Kalman filter, the detection of attacks is judged according to the residual value of the system. According to the disruption resources and detectability, a general security risk level model is further established to evaluate the security risk level of the system under attacks. The simulation experiments are conducted by using Matlab to analyze the destructiveness and detectability of attacks, where the results show that the proposed qualitative analysis method can effectively describe the security risk under the cyber-attacks.

This paper is concerned with the problem of adaptive finite-time prescribed performance control for a category of uncertain nonlinear systems subject to fuzzy dead-zone input. Via combining the technologies of command filter and backstepping control, the ``singularity'' and the ``explosion of complexity'' issues within controller design procedure are avoided. Moreover, by designing a state observer and utilizing the center-of-gravity theorem, the unmeasured states of system are estimated and the fuzzy issue result from fuzzy dead-zone input is disposed, respectively. Meanwhile, a finite-time fuzzy controller is constructed via combining with finite-time stability criterion, which guarantees all the signals in closed-loop system are convergent and the trajectory of tracking error also strictly evolves within a predefined range in finite time. At last, some simulation results confirm the viability of presented theoretical results.

Adaptive feedback cancellation (AFC) method is widely adopted for the purpose of reducing the adverse effects of acoustic feedback on the sound reinforcement systems. However, since the existence of forward path results in the correlation between the source signal and the feedback signal, the source signal is mistakenly considered as the feedback signal to be eliminated by adaptive filter when it is colored, which leads to a inaccurate prediction of the acoustic feedback signal. In order to solve this problem, prediction error method is introduced in this paper to remove the correlation between the source signal and the feedback signal. Aiming at the dilemma of Modified Least Mean Square (MLMS) algorithm in choosing between prediction speed and prediction accuracy, an improved MLMS algorithm with a variable step-size scheme is proposed. Simulation examples are applied to show that the proposed algorithm can obtain more accurate prediction of acoustic feedback signal in a shorter time than the MLMS algorithm.

Over the past few years computer vision has become the essential aspect of modern era of technology. This computer vision is manly based on image processing whereas the image processing includes three important aspects as image filtering, image compression & image security. The image filtering can be achieved by using various filtering techniques but the PSNR & operating frequency are the most challenging aspects of image filtering. This paper mainly focused on overcoming the challenges appears while removing the salt & pepper noise with conventional median filtering by developing improved adaptive moving window architecture median filter & comparing its performance to have improved performance in terms of PSNR & operating frequency.

Cyber-Physical Systems (CPS) will form the basis for the world's critical infrastructure and, thus, have the potential to significantly impact human lives in the near future. In recent years, there has been an increasing demand for connectivity in CPS, which has brought to attention the issue of cyber security. Aside from traditional information systems threats, CPS faces new challenges due to the heterogeneity of devices and protocols. In this paper, we investigate how Feature Selection may improve intrusion detection accuracy. In particular, we propose an adapted Greedy Randomized Adaptive Search Procedure (GRASP) metaheuristic to improve the classification performance in CPS perception layer. Our numerical results reveal that GRASP metaheuristic overcomes traditional filter-based feature selection methods for detecting four attack classes in CPSs.

We investigate if the random feature selection approach proposed in [1] to improve the robustness of forensic detectors to targeted attacks, can be extended to detectors based on deep learning features. In particular, we study the transferability of adversarial examples targeting an original CNN image manipulation detector to other detectors (a fully connected neural network and a linear SVM) that rely on a random subset of the features extracted from the flatten layer of the original network. The results we got by considering three image manipulation detection tasks (resizing, median filtering and adaptive histogram equalization), two original network architectures and three classes of attacks, show that feature randomization helps to hinder attack transferability, even if, in some cases, simply changing the architecture of the detector, or even retraining the detector is enough to prevent the transferability of the attacks.

Steganalysis is an interesting classification problem in order to discriminate the images, including hidden messages from the clean ones. There are many methods, including deep CNN networks to extract fine features for this classification task. Nevertheless, a few researches have been conducted to improve the final classifier. Some state-of-the-art methods try to ensemble the networks by a voting strategy to achieve more stable performance. In this paper, a selection phase is proposed to filter improper networks before any voting. This filtering is done by a binary relevance multi-label classification approach. The Logistic Regression (LR) is chosen here as the last layer of network for classification. The large-margin Fisher’s linear discriminant (FLD) classifier is assigned to each one of the networks. It learns to discriminate the training instances which associated network is suitable for or not. Xu-Net, one of the most famous state-of-the-art Steganalysis models, is chosen as the base networks. The proposed method with different approaches is applied on the BOSSbase dataset and is compared with traditional voting and also some state-of-the-art related ensemble techniques. The results show significant accuracy improvement of the proposed method in comparison with others.

Recently, the field of adversarial machine learning has been garnering attention by showing that state-of-the-art deep neural networks are vulnerable to adversarial examples, stemming from small perturbations being added to the input image. Adversarial examples are generated by a malicious adversary by obtaining access to the model parameters, such as gradient information, to alter the input or by attacking a substitute model and transferring those malicious examples over to attack the victim model. Specifically, one of these attack algorithms, Robust Physical Perturbations (RP2), generates adversarial images of stop signs with black and white stickers to achieve high targeted misclassification rates against standard-architecture traffic sign classifiers. In this paper, we propose BlurNet, a defense against the RP2 attack. First, we motivate the defense with a frequency analysis of the first layer feature maps of the network on the LISA dataset, which shows that high frequency noise is introduced into the input image by the RP2 algorithm. To remove the high frequency noise, we introduce a depthwise convolution layer of standard blur kernels after the first layer. We perform a blackbox transfer attack to show that low-pass filtering the feature maps is more beneficial than filtering the input. We then present various regularization schemes to incorporate this lowpass filtering behavior into the training regime of the network and perform white-box attacks. We conclude with an adaptive attack evaluation to show that the success rate of the attack drops from 90% to 20% with total variation regularization, one of the proposed defenses.

This paper proposes a method for detecting anomalies in video data. A Variational Autoencoder (VAE) is used for reducing the dimensionality of video frames, generating latent space information that is comparable to low-dimensional sensory data (e.g., positioning, steering angle), making feasible the development of a consistent multi-modal architecture for autonomous vehicles. An Adapted Markov Jump Particle Filter defined by discrete and continuous inference levels is employed to predict the following frames and detecting anomalies in new video sequences. Our method is evaluated on different video scenarios where a semi-autonomous vehicle performs a set of tasks in a closed environment.

Global Navigation Satellite System (GNSS) jamming is an evolving technology where new modulations are progressively introduced in order to reduce the impact of interference mitigation techniques such as Adaptive Notch Filters (ANFs). The Standardisation of GNSS Threat reporting and Receiver testing through International Knowledge Exchange, Experimentation and Exploitation (STRIKE3) project recently described a new class of jamming signals, called tick signals, where a basic frequency tick is hopped over a large frequency range. In this way, discontinuities are introduced in the instantaneous frequency of the jamming signals. These discontinuities reduce the effectiveness of ANFs, which unable to track the jamming signal. This paper analyses the effectiveness of interference mitigation techniques with respect to frequency-hopped tick jamming signals. ANFs and Robust Interference Mitigation (RIM) techniques are analysed. From the analysis, it emerges that, despite the presence of frequency discontinuities, ANFs provide some margin against tick signals. However, frequency discontinuities prevent ANFs to remove all the jamming components and receiver operations are denied for moderate Jamming to Noise power ratio (J/N) values, RIM techniques are not affected by the presence of frequency discontinuities and significantly higher jamming power are sustained by the receiver when this type of techniques is adopted.

Publicly available blacklists are popular tools to capture and spread information about misbehaving entities on the Internet. In some cases, their straight-forward utilization leads to many false positives. In this work, we propose a system that combines blacklists with network flow data while introducing automated evaluation techniques to avoid reporting unreliable alerts. The core of the system is formed by an Adaptive Filter together with an Evaluator module. The assessment of the system was performed on data obtained from a national backbone network. The results show the contribution of such a system to the reduction of unreliable alerts.

A recurring principle in consideration of the future of systems engineering is continual dynamic adaptation. Context drives change whether it be from potential loss (threats, vulnerabilities) or from potential gain (opportunity-driven). Contextual-awareness has great influence over the future of systems engineering and of systems security. Those contextual environments contain fitness functions that will naturally select compatible approaches and filter out the incompatible, with prejudice. We don't have to guess at what those environmental shaping forces will look like. William Gibson famously tells us why: “The future is already here, it's just not evenly distributed;” and, sometimes difficult to discern. This paper provides archetypes that 1) characterize general systems engineering for products, processes, and operations; 2) characterize the integration of security to systems engineering; and, 3) characterize contextually aware agile-security. This paper is more of a problem statement than a solution. Solution objectives and tactics for guiding the path forward have a broader range of options for subsequent treatment elsewhere. Our purpose here is to offer a short list of necessary considerations for effective contextually aware adaptive system security in the future of systems engineering.

{The paper considers the efficiency of an adaptive non-recursive filter using the adjustment algorithm for weighting coefficients taking into account the constant envelope of the desired signal when receiving signals with multi-position phase shift keying against the background of noise and non-fluctuation interference. Two types of such interference are considered - harmonic and retranslated. The optimal filter parameters (adaptation coefficient and length) are determined by using simulation; the effect of the filter on the noise immunity of a quadrature coherent signal receiver with multi-position phase shift keying for different combinations of interference and their intensity is estimated. It is shown that such an adaptive filter can successfully deal with the most dangerous sighting harmonic interference}.

With the ever-increasing number of smart home devices, the issues related to these environments are also growing. With an ever-growing attack surface, there is no standard way to protect homes and their inhabitants from new threats. The inhabitants are rarely aware of the increased security threats that they are exposed to and how to manage them. To tackle this problem, we propose a solution based on segmented architectures similar to the ones used in industrial systems. In this approach, the smart home is segmented into various levels, which can broadly be categorised into an inner level and external level. The external level is protected by a firewall that checks the communication from/to the Internet to/from the external devices. The internal level is protected by an additional firewall that filters the information and the communications between the external and the internal devices. This segmentation guarantees a trusted environment among the entities of the internal network. In this paper, we propose an adaptive trust model that checks the behaviour of the entities and in case the entities violate trust rules they can be put in quarantine or banned from the network.

This paper considers consensus-based distributed Kalman filtering subject to data-falsification attack, where Byzantine agents share manipulated data with their neighboring agents. The attack is assumed to be coordinated among the Byzantine agents and follows a linear model. The goal of the Byzantine agents is to maximize the network-wide estimation error while evading false-data detectors at honest agents. To that end, we propose a joint selection of Byzantine agents and covariance matrices of attack sequences to maximize the network-wide estimation error subject to constraints on stealthiness and the number of Byzantine agents. The attack strategy is then obtained by employing block-coordinate descent method via Boolean relaxation and backward stepwise based subset selection method. Numerical results show the efficiency of the proposed attack strategy in comparison with other naive and uncoordinated attacks.

Embedding costs used in content-adaptive image steganographic schemes can be defined in a heuristic way or with a statistical model. Inspired by previous steganographic methods, i.e., MG (multivariate Gaussian model) and MiPOD (minimizing the power of optimal detector), we propose a model-driven scheme in this paper. Firstly, we model image residuals obtained by high-pass filtering with quantized multivariate Gaussian distribution. Then, we derive the approximated Fisher Information (FI). We show that FI is related to both Gaussian variance and filter coefficients. Lastly, by selecting the maximum FI value derived with various filters as the final FI, we obtain embedding costs. Experimental results show that the proposed scheme is comparable to existing steganographic methods in resisting steganalysis equipped with rich models and selection-channel-aware rich models. It is also computational efficient when compared to MiPOD, which is the state-of-the-art model-driven method.

Modern cyber-physical systems are increasingly complex and vulnerable to attacks like false data injection aimed at destabilizing and confusing the systems. We develop and evaluate an attack-detection framework aimed at learning a dynamic invariant network, data-driven temporal causal relationships between components of cyber-physical systems. We evaluate the relative performance in attack detection of the proposed model relative to traditional anomaly detection approaches. In this paper, we introduce Granger Causality based Kalman Filter with Adaptive Robust Thresholding (G-KART) as a framework for anomaly detection based on data-driven functional relationships between components in cyber-physical systems. In particular, we select power systems as a critical infrastructure with complex cyber-physical systems whose protection is an essential facet of national security. The system presented is capable of learning with or without network topology the task of detection of false data injection attacks in power systems. Kalman filters are used to learn and update the dynamic state of each component in the power system and in-turn monitor the component for malicious activity. The ego network for each node in the invariant graph is treated as an ensemble model of Kalman filters, each of which captures a subset of the node's interactions with other parts of the network. We finally also introduce an alerting mechanism to surface alerts about compromised nodes.

Large scale cloud networks consist of distributed networking and computing elements that process critical information and thus security is a key requirement for any environment. Unfortunately, assessing the security state of such networks is a challenging task and the tools used in the past by security experts such as packet filtering, firewall, Intrusion Detection Systems (IDS) etc., provide a reactive security mechanism. In this paper, we introduce a Moving Target Defense (MTD) based proactive security framework for monitoring attacks which lets us identify and reason about multi-stage attacks that target software vulnerabilities present in a cloud network. We formulate the multi-stage attack scenario as a two-player zero-sum Markov Game (between the attacker and the network administrator) on attack graphs. The rewards and transition probabilities are obtained by leveraging the expert knowledge present in the Common Vulnerability Scoring System (CVSS). Our framework identifies an attacker's optimal policy and places countermeasures to ensure that this attack policy is always detected, thus forcing the attacker to use a sub-optimal policy with higher cost.

In this paper, we design NETHCF, a line-rate in-network system for filtering spoofed traffic. NETHCF leverages the opportunity provided by programmable switches to design a novel defense against spoofed IP traffic, and it is highly efficient and adaptive. One key challenge stems from the restrictions of the computational model and memory resources of programmable switches. We address this by decomposing the HCF system into two complementary components-one component for the data plane and another for the control plane. We also aggregate the IP-to-Hop-Count (IP2HC) mapping table for efficient memory usage, and design adaptive mechanisms to handle end-to-end routing changes, IP popularity changes, and network activity dynamics. We have built a prototype on a hardware Tofino switch, and our evaluation demonstrates that NETHCF can achieve line-rate and adaptive traffic filtering with low overheads.