Search

In this work, we consider transferring the structure information from large networks to compact ones for dense prediction tasks in computer vision. Previous knowledge distillation strategies used for dense prediction tasks often directly borrow the distillation scheme for image classification and perform knowledge distillation for each pixel separately, leading to sub-optimal performance. Here we propose to distill structured knowledge from large networks to compact networks, taking into account the fact that dense prediction is a structured prediction problem. Specifically, we study two structured distillation schemes: i) pair-wise distillation that distills the pair-wise similarities by building a static graph; and ii) holistic distillation that uses adversarial training to distill holistic knowledge. The effectiveness of our knowledge distillation approaches is demonstrated by experiments on three dense prediction tasks: semantic segmentation, depth estimation and object detection. Code is available at: https://git.io/StructKD, Comment: v1:10 pages cvpr2019 accepted; v2:15 pages for a journal version; Code is available at: https://github.com/irfanICMLL/structure_knowledge_distillation; fix typos

Nearest neighbor search aims to obtain the samples in the database with the smallest distances from them to the queries, which is a basic task in a range of fields, including computer vision and data mining. Hashing is one of the most widely used methods for its computational and storage efficiency. With the development of deep learning, deep hashing methods show more advantages than traditional methods. In this survey, we detailedly investigate current deep hashing algorithms including deep supervised hashing and deep unsupervised hashing. Specifically, we categorize deep supervised hashing methods into pairwise methods, ranking-based methods, pointwise methods as well as quantization according to how measuring the similarities of the learned hash codes. Moreover, deep unsupervised hashing is categorized into similarity reconstruction-based methods, pseudo-label-based methods and prediction-free self-supervised learning-based methods based on their semantic learning manners. We also introduce three related important topics including semi-supervised deep hashing, domain adaption deep hashing and multi-modal deep hashing. Meanwhile, we present some commonly used public datasets and the scheme to measure the performance of deep hashing algorithms. Finally, we discuss some potential research directions in conclusion., Comment: Accepted by ACM Transactions on Knowledge Discovery from Data (TKDD)

This paper evaluates outage probability (OP) and intercept probability (IP) of physical-layer security based MIMO networks adopting cooperative jamming (Coop-Jam). In the considered scenario, a multi-antenna source communicates with a multi-antenna destination employing transmit antenna selection (TAS)/ selection combining (SC), in presence of a multi-antenna eavesdropper using SC. One of jammers appearing near the destination is selected for generating jamming noises on the eavesdropper. Moreover, the destination supports the wireless energy for the chosen jammer, and cooperates with it to remove the jamming noises. We consider two jammer selection approaches, named RAND and SHORT. In RAND, the destination randomly selects the jammer, and in SHORT, the jammer, which is nearest to the destination, is chosen. We derive exact and asymptotic expressions of OP and IP over Rayleigh fading, and perform Monte-Carlo simulations to verify the correction of our derivation. The results present advantages of the proposed RAND and SHORT methods, as compared with the corresponding one without using Coop-Jam.

Vertex leaky forcing was recently introduced as a new variation of zero forcing in order to show how vertex leaks can disrupt the zero forcing process in a graph. An edge leak is an edge that is not allowed to be forced across during the zero forcing process. The $\ell$-edge-leaky forcing number of a graph is the size of a smallest zero forcing set that can force the graph blue despite $\ell$ edge leaks. This paper contains an analysis of the effect of edge leaks on the zero forcing process instead of vertex leaks. Furthermore, specified $\ell$-leaky forcing is introduced. The main result is that $\ell$-leaky forcing, $\ell$-edge-leaky forcing, and specified $\ell$-leaky forcing are equivalent. Furthermore, all of these different kinds of leaks can be mixed so that vertex leaks, edge leaks, and specified leaks are used. This mixed $\ell$-leaky forcing number is also the same as the (vertex) $\ell$-leaky forcing number., Comment: arXiv admin note: text overlap with arXiv:2008.06552

Protecting the parameters, states, and input/output signals of a dynamic controller is essential for securely outsourcing its computation to an untrusted third party. Although a fully homomorphic encryption scheme allows the evaluation of controller operations with encrypted data, an encrypted dynamic controller with the encryption scheme destabilizes a closed-loop system or degrades the control performance due to overflow. This paper presents a novel controller representation based on input-output history data to implement an encrypted dynamic controller that operates without destabilization and performance degradation. Implementation of this encrypted dynamic controller representation can be optimized via batching techniques to reduce the time and space complexities. Furthermore, this study analyzes the stability and performance degradation of a closed-loop system caused by the effects of controller encryption. A numerical simulation demonstrates the feasibility of the proposed encrypted control scheme, which inherits the control performance of the original controller at a sufficient level., Comment: 12 pages, 7 figures

Average consensus is extensively used in distributed networks for computation and control, where all the agents constantly communicate with each other and update their states in order to reach an agreement. Under a general average consensus algorithm, information exchanged through wireless or wired communication networks could lead to the disclosure of sensitive and private information. In this paper, we propose a privacy-preserving push-sum approach for directed networks that can protect the privacy of all agents while achieving average consensus simultaneously. Each node decomposes its initial state arbitrarily into two substates, and their average equals to the initial state, guaranteeing that the agent's state will converge to the accurate average consensus. Only one substate is exchanged by the node with its neighbours over time, and the other one is reserved. That is to say, only the exchanged substate would be visible to an adversary, preventing the initial state information from leakage. Different from the existing state-decomposition approach which only applies to undirected graphs, our proposed approach is applicable to strongly connected digraphs. In addition, in direct contrast to offset-adding based privacy-preserving push-sum algorithm, which is vulnerable to an external eavesdropper, our proposed approach can ensure privacy against both an honest-but-curious node and an external eavesdropper. A numerical simulation is provided to illustrate the effectiveness of the proposed approach.

This paper deals with the state estimation problem in discrete-event systems modeled with nondeterministic finite automata, partially observed via a sensor measuring unit whose measurements (reported observations) may be vitiated by a malicious attacker. The attacks considered in this paper include arbitrary deletions, insertions, or substitutions of observed symbols by taking into account a bounded number of attacks or, more generally, a total cost constraint (assuming that each deletion, insertion, or substitution bears a positive cost to the attacker). An efficient approach is proposed to describe possible sequences of observations that match the one received by the measuring unit, as well as their corresponding state estimates and associated total costs. We develop an algorithm to obtain the least-cost matching sequences by reconstructing only a finite number of possible sequences, which we subsequently use to efficiently perform state estimation. We also develop a technique for verifying tamper-tolerant diagnosability under attacks that involve a bounded number of deletions, insertions, and substitutions (or, more generally, under attacks of bounded total cost) by using a novel structure obtained by attaching attacks and costs to the original plant. The overall construction and verification procedure have complexity that is of O(|X|^2 C^2),where |X| is the number of states of the given finite automaton and C is the maximum total cost that is allowed for all the deletions, insertions, and substitutions. We determine the minimum value of C such that the attacker can coordinate its tampering action to keep the observer indefinitely confused while utilizing a finite number of attacks. Several examples are presented to demonstrate the proposed methods.

Recent developments in communication technologies, such as 5G, together with innovative computing paradigms, such as edge computing, provide further possibilities for the implementation of real-time networked control systems. However, privacy and cyber-security concerns arise when sharing private data between sensors, agents and a third-party computing facility. In this paper, a secure version of the distributed formation control is presented, analyzed and simulated, where gradient-based formation control law is implemented in the edge, with sensor and actuator information being secured by fully homomorphic encryption method based on learning with error (FHE-LWE) combined with a proposed mixed uniform-logarithmic quantizer (MULQ). The novel quantizer is shown to be suitable for realizing secure control systems with FHE-LWE where the critical real-time information can be quantized into a prescribed bounded space of plaintext while satisfying a sector bound condition whose lower and upper-bound can be made sufficiently close to an identity. An absolute stability analysis is presented, that shows the asymptotic stability of the closed-loop secure control system.

A bus-off attack is a denial-of-service (DoS) attack which exploits error handling in the controller area network (CAN) to induce an honest node to disconnect itself from the CAN bus. This paper develops a stochastic transmission policy as a countermeasure for the controller-transmitter pair against the bus-off attack. We model this as a non-zero-sum linear-quadratic-Gaussian game between the controller-transmitter pair and the attacker. We derive Nash equilibria of the game for two different information structures of the attacker. We show that the attacker has a dominant attack strategy under both information structures. Under the dominant attack strategy, we show that the optimal control policy is linear in the system state. We further identify a necessary and a sufficient conditions on the transmission policy to have bounded average cost. The theoretical results are complemented by a detailed case study of a bus-off attack on a vehicular adaptive cruise control model.

We study privacy-utility trade-offs where users share privacy-correlated useful information with a service provider to obtain some utility. The service provider is adversarial in the sense that it can infer the users' private information based on the shared useful information. To minimize the privacy leakage while maintaining a desired level of utility, the users carefully perturb the useful information via a probabilistic privacy mapping before sharing it. We focus on the setting in which the adversary attempting an inference attack on the users' privacy has potentially biased information about the statistical correlation between the private and useful variables. This information asymmetry between the users and the limited adversary leads to better privacy guarantees than the case of the omniscient adversary under the same utility requirement. We first identify assumptions on the adversary's information so that the inference costs are well-defined and finite. Then, we characterize the impact of the information asymmetry and show that it increases the inference costs for the adversary. We further formulate the design of the privacy mapping against a limited adversary using a difference of convex functions program and solve it via the concave-convex procedure. When the adversary's information is not precisely available, we adopt a Bayesian view and represent the adversary's information by a probability distribution. In this case, the expected cost for the adversary does not admit a closed-form expression, and we establish and maximize a lower bound of the expected cost. We provide a numerical example regarding a census data set to illustrate the theoretical results.

We consider the problem of secure packet routing at the maximum achievable rate in Quantum Key Distribution (QKD) networks. Assume that a QKD protocol generates symmetric private key pairs for secure communication over each link in a network. The quantum key generation process is modeled using a stochastic counting process. Packets are first encrypted with the quantum keys available for each hop and then transmitted on a point-to-point basis over the links. A fundamental problem in this setting is the design of a secure and capacity-achieving routing policy that takes into account the time-varying availability of the encryption keys and finite link capacities. In this paper, we propose a new secure throughput-optimal policy called Tandem Queue Decomposition (TQD). The TQD policy is derived by combining the QKD process with the Universal Max Weight routing policy, proposed earlier by Sinha and Modiano. We show that the TQD policy solves the problem of secure and efficient packet routing for a broad class of traffic, including unicast, broadcast, multicast, and anycast. The proposed decomposition reduces the problem to the generalized network flow problem without the key availability constraints over a transformed network. The proof of the throughput-optimality of the TQD policy uses the Lyapunov stability theory for analyzing the interdependent packet queueing process and the key-storage dynamics. Finally, we demonstrate the competitiveness of the TQD policy over the existing algorithms by numerically comparing them on a simulator that we build on top of the state-of-the-art OMNeT++ network simulator platform.

We study the error of linear regression in the face of adversarial attacks. In this framework, an adversary changes the input to the regression model in order to maximize the prediction error. We provide bounds on the prediction error in the presence of an adversary as a function of the parameter norm and the error in the absence of such an adversary. We show how these bounds make it possible to study the adversarial error using analysis from non-adversarial setups. The obtained results shed light on the robustness of overparameterized linear models to adversarial attacks. Adding features might be either a source of additional robustness or brittleness. On the one hand, we use asymptotic results to illustrate how double-descent curves can be obtained for the adversarial error. On the other hand, we derive conditions under which the adversarial error can grow to infinity as more features are added, while at the same time, the test error goes to zero. We show this behavior is caused by the fact that the norm of the parameter vector grows with the number of features. It is also established that $\ell_\infty$ and $\ell_2$-adversarial attacks might behave fundamentally differently due to how the $\ell_1$ and $\ell_2$-norms of random projections concentrate. We also show how our reformulation allows for solving adversarial training as a convex optimization problem. This fact is then exploited to establish similarities between adversarial training and parameter-shrinking methods and to study how the training might affect the robustness of the estimated models.

In the context of digital signatures, the proxy signature holds a significant role of enabling an original signer to delegate its signing ability to another party (i.e., proxy signer). It has significant practical applications. Particularly it is useful in distributed systems, where delegation of authentication rights is quite common. For example, key sharing protocol, grid computing, and mobile communications. Currently, a large portion of existing proxy signature schemes are based on the hardness of problems like integer factoring, discrete logarithms, and/or elliptic curve discrete logarithms. However, with the rising of quantum computers, the problem of prime factorization and discrete logarithm will be solvable in polynomial-time, due to Shor's algorithm, which dilutes the security features of existing ElGamal, RSA, ECC, and the proxy signature schemes based on these problems. As a consequence, construction of secure and efficient post-quantum proxy signature becomes necessary. In this work, we develop a post-quantum proxy signature scheme Mult-proxy, relying on multivariate public key cryptography (MPKC), which is one of the most promising candidates of post-quantum cryptography. We employ a 5-pass identification protocol to design our proxy signature scheme. Our work attains the usual proxy criterion and a one-more-unforgeability criterion under the hardness of the Multivariate Quadratic polynomial (MQ) problem. It produces optimal size proxy signatures and optimal size proxy shares in the field of MPKC.

This paper presents a novel image encryption scheme based on extended large-scale randomization key-stream generator. The basic form of the key-stream generator is presented, and employed in digital image ciphering. The modification of the basic form also, presented, and gives encouraging results in image encryption as compared with classical non-linear stream cipher generators and the basic form. Pixel shuffling is performed via vertical and horizontal permutation. Shuffling is used to expand diffusion in the image and dissipate high correlation among image pixels the sequences generated from all presented generators are introduced to five well-known statistical tests of randomness to judge their randomness characteristic. The ciphered images are tested for their residual intelligibility subjectively. The measures applied to images ciphered by one of the classical key-stream cipher generators (Threshold generator) for the purpose of comparison with the presented key-stream algorithms. Experiments results show that the proposed algorithm achieves the image security. In order to evaluate performance, the proposed algorithm was measured through a series of tests. Experimental results illustrate that the proposed scheme shows a good resistance against brute-force and statistical attacks.

This paper is concerned with the security issue in the state estimation problem for a networked control system (NCS). A new model of joint false data injection (FDI) attack is established wherein attacks are injected to both the remote estimator and the communication channels. Such a model is general that includes most existing FDI attack models as special cases. The joint FDI attacks are subjected to limited access and/or resource constraints, and this gives rise to a few attack scenarios to be examined one by one. Our objective is to establish the so-called insecurity conditions under which there exists an attack sequence capable of driving the estimation bias to infinity while bypassing the anomaly detector. By resorting to the generalized inverse theory, necessary and sufficient conditions are derived for the insecurity under different attack scenarios. Subsequently, easy-to-implement algorithms are proposed to generate attack sequences on insecure NCSs with respect to different attack scenarios. In particular, by using a matrix splitting technique, the constraint-induced sparsity of the attack vectors is dedicatedly investigated. Several numerical examples are

We show that the most common flavors of noisy leakage can be simulated in the information-theoretic setting using a single query of bounded leakage, up to a small statistical simulation error and a slight loss in the leakage parameter. The latter holds true in particular for one of the most used noisy-leakage models, where the noisiness is measured using the conditional average min-entropy (Naor and Segev, CRYPTO'09 and SICOMP'12). Our reductions between noisy and bounded leakage are achieved in two steps. First, we put forward a new leakage model (dubbed the dense leakage model) and prove that dense leakage can be simulated in the information-theoretic setting using a single query of bounded leakage, up to small statistical distance. Second, we show that the most common noisy-leakage models fall within the class of dense leakage, with good parameters. Third, we prove lower bounds on the amount of bounded leakage required for simulation with sub-constant error, showing that our reductions are nearly optimal. In particular, our results imply that useful general simulation of noisy leakage based on statistical distance and mutual information is impossible. We also provide a complete picture of the relationships between different noisy-leakage models. Our result finds applications to leakage-resilient cryptography, where we are often able to lift security in the presence of bounded leakage to security in the presence of noisy leakage, both in the information-theoretic and in the computational setting. Remarkably, this lifting procedure makes only black-box use of the underlying schemes. Additionally, we show how to use lower bounds in communication complexity to prove that bounded-collusion protocols (Kumar, Meka, and Sahai, FOCS'19) for certain functions do not only require long transcripts, but also necessarily need to reveal enough information about the inputs.

Motivated by recent security issues in cyber-physical systems, this technical note studies the stabilization problem of networked control systems under Denial-of-Service (DoS) attacks. In particular, we consider to stabilize a nonlinear system with limited data rate via linearization. We employ a deterministic DoS attack model constrained in terms of attacks' frequency and duration, allowing us to cover a large class of potential attacks. To achieve asymptotic stabilization, we propose a resilient dynamic quantizer in the sense that it does not saturate in the presence of packet losses caused by DoS attacks. A sufficient condition for stability is derived by restricting the average DoS frequency and duration. In addition, because of the locality of linearization, we explicitly investigate an estimate of the region of attraction, which can be expected to be reduced depending on the strength of DoS attacks. A simulation example is presented for demonstration of our results.

International audience; The Goppa Code Distinguishing (GD) problem asks to distinguish efficiently a generator matrix of a Goppa code from a randomly drawn one. We revisit a distinguisher for alternant and Goppa codes through a new approach, namely by studying the dimension of square codes. We provide here a rigorous upper bound for the dimension of the square of the dual of an alternant or Goppa code, while the previous approach only provided algebraic explanations based on heuristics. Moreover, for Goppa codes, our proof extends to the non-binary case as well, thus providing an algebraic explanation for the distinguisher which was missing up to now. All the upper bounds are tight and match experimental evidence. Our work also introduces new algebraic results about products of trace codes in general and of dual of alternant and Goppa codes in particular, clarifying their square code structure. This might be of interest for cryptanalysis purposes.

This article proposes a model predictive control as a secure service (MPCaaSS) framework for cyber-physical systems (CPS) in the presence of both cyber threats and external disturbances. Firstly, in order to take advantage of the cloud-edge computing, we design a double-layer controller architecture by using a novel control parameterization based on Gaussian radial basis functions. In this controller architecture, the cloud-side controller optimizes the controller parameters of the edge-side controller, whereas the edge-side controller implements the real-time control law using the generated controller parameters. Secondly, in order to securely transmit data packets, we integrate an encoding scheme and an elliptic curve cryptography (ECC) based encryption into the proposed MPCaaSS framework. Then, the controller parameters and the state measurements can be encrypted such that no malicious attackers can corrupt and intercept the transmission. It is shown that the recursive feasibility of MPCaaSS is achieved under some sufficient conditions, and the robust stability of the closed-loop system is guaranteed if the optimization problem is recursively feasible. Simulated examples are conducted to demonstrate the effectiveness of the proposed method.

With the wide application of cyber-physical systems, stealthy attacks on remote state estimation have attracted increasing research attention. Recently, various stealthy innovation-based linear attack models were proposed, in which the relaxed stealthiness was measured by the Kullback--Leibler divergence. This paper studies existing innovation-based linear attack strategies with relaxed stealthiness and concludes that all of them provided merely suboptimal solutions. The main reason is some oversight in solving the involved optimization problems: some covariance constraints were not perfectly handled. This paper provides the corresponding optimal solutions for those stealthy attacks. Both one-step and holistic optimizations of stealthy attacks are studied, and the worst-case attacks with and without zero-mean constraints are derived analytically, without the necessity to formulate semidefinite programming problems.

In this paper, we investigate secure transmissions in integrated satellite-terrestrial communications and the green interference based symbiotic security scheme is proposed. Particularly, the co-channel interference induced by the spectrum sharing between satellite and terrestrial networks and the inter-beam interference due to frequency reuse among satellite multi-beam serve as the green interference to assist the symbiotic secure transmission, where the secure transmissions of both satellite and terrestrial links are guaranteed simultaneously. Specifically, to realize the symbiotic security, we formulate a problem to maximize the sum secrecy rate of satellite users by cooperatively beamforming optimizing and a constraint of secrecy rate of each terrestrial user is guaranteed. Since the formulated problem is non-convex and intractable, the Taylor expansion and semi-definite relaxation (SDR) are adopted to further reformulate this problem, and the successive convex approximation (SCA) algorithm is designed to solve it. Finally, the tightness of the relaxation is proved. In addition, numerical results verify the efficiency of our proposed approach.

A network of observers is considered, where through asynchronous (with bounded delay) communications, they cooperatively estimate the states of a Linear Time-Invariant (LTI) system. In such a setting, a new type of adversary might affect the observation process by impersonating the identity of the regular node, which is a violation of communication authenticity. These adversaries also inherit the capabilities of Byzantine nodes, making them more powerful threats called smart spoofers. We show how asynchronous networks are vulnerable to smart spoofing attack. In the estimation scheme considered in this paper, information flows from the sets of source nodes, which can detect a portion of the state variables each, to the other follower nodes. The regular nodes, to avoid being misguided by the threats, distributively filter the extreme values received from the nodes in their neighborhood. Topological conditions based on strong robustness are proposed to guarantee the convergence. Two simulation scenarios are provided to verify the results.

Homomorphic Proxy Re-Encryption (HPRE) is an extension of Proxy Re-Encryption (PRE) which combines the advantages of both Homomorphic Encryption (HE) and PRE. A HPRE scheme allows arbitrary evaluations to be performed on ciphertexts under one (the delegator's) public key and, using a re-encryption key, it transforms the resulting ciphertext to a new ciphertext under another (the delegatee's) public key. Prior HPRE schemes are either CPA-secure or CCA-secure but only support partial homomorphic operations. We propose a generic construction of single-hop HPRE scheme which supports fully homomorphic operations. The proposed scheme is proven secure in our new index-based CCA-HPRE model. Our technique is to give a generic transformation that turns any multi-identity identity-based FHE (IBFHE) scheme with key switching into Fully Homomorphic Encryption (FHE) with key switching from which we can obtain the proposed single-hop HPRE scheme. We also present a concrete instantiation of multi-identity IBFHE with key switching from learning with errors (LWE) in the standard model.

Machine learning (ML) classification is increasingly used in safety-critical systems. Protecting ML classifiers from adversarial examples is crucial. We propose that the main threat is that of an attacker perturbing a confidently classified input to produce a confident misclassification. To protect against this we devise an adversarial bound (AB) for a Gaussian process classifier, that holds for the entire input domain, bounding the potential for any future adversarial method to cause such misclassification. This is a formal guarantee of robustness, not just an empirically derived result. We investigate how to configure the classifier to maximise the bound, including the use of a sparse approximation, leading to the method producing a practical, useful and provably robust classifier, which we test using a variety of datasets., Comment: 10 pages + 2 pages references + 7 pages of supplementary. 12 figures. Submitted to AAAI

In this paper, a robust Chaos Based Stream Cipher (CBSC) is proposed . The novelty of this work is that it addresses all challenges confronting chaos- based cryptography. The PCBSC (Proposed CBSC) has a robust synchronization circuit that mitigates the effect of channel noise, a perturbation block that overcomes the dynamical degradation, a robust encryption scheme, and an efficient control parameters ' generator that generates strong keys. According to the complexity evaluation , the improved chaotic map provides good statistical properties. This can be confirmed by the obtained high values of the statistical metrics (Largest Lyapunov Exponent, Approximate Entropy, Permutation Entropy, and Sample Entropy) used for the evaluation . According to the security analysis ; the PCBSC has good security features , the PCBSC provides strong keys that ensure confusion property, as well as enough space to withstand brute-force attacks . On the other hand , the proposed encryption scheme proves its efficiency; the result of the differential attack clearly shows that the diffusion property is guaranteed. Additionally, the original images' statistical properties are completely dispersed on the encrypted images. The obtained performance over noisy channels proves the synchronization circuit's efficiency. When compared to other proposals, the PCBSC provides the best results. In addition, the PCBSC is implemented on an FPGA and evaluated in real-time over a wireless link.

This article is concerned with the secure state estimation problem for cyber-physical systems under intermittent denial-of-service (DoS) attacks. Based on a switching scheme and the cascade observer technique, a novel resilient state observer with a switched compensation mechanism is designed. Moreover, a quantitative relationship between the resilience against DoS attacks and the design parameters is revealed. Compared with the existing results, where only the boundedness of the estimation error is guaranteed under DoS attacks, the exponential convergence of the estimation error is achieved by employing the proposed observer scheme, such that the estimation performance is improved. More specifically, in the disturbance-free case, it is proven that the state estimation error converges exponentially to 0 despite the existence of DoS attacks. Finally, simulation results are provided to illustrate the effectiveness and merits of the proposed methods.

It has been debated whether protective measurement implies the reality of the wave function. In this paper, I present a new analysis of the relationship between protective measurements and the reality of the wave function. First, I briefly introduce protective measurements and the ontological models framework for them. Second, I give a simple proof of Hardy's theorem in terms of protective measurements. Third, I analyze two suggested $\psi$-epistemic models of a protective measurement. It is shown that although these models can explain the appearance of expectation values of observables in a single measurement, their predictions about the variance of the result of a non-ideal protective measurement are different from those of quantum mechanics. Finally, I argue that under an auxiliary finiteness assumption about the dynamics of the ontic state, protective measurement implies the reality of the wave function in the ontological models framework., Comment: 25 pages, no figures. arXiv admin note: text overlap with arXiv:1611.02738

The problem of secure finite-horizon consensus control for discrete time-varying multiagent systems (MASs) with actuator saturation and cyber attacks is addressed in this article. A random attack model is first proposed to account for randomly occurring false data injection attacks and denial-of-service attacks, whose dynamics are governed by the random Markov process. The hybrid secure control scheme is developed to mitigate the influence of arbitrary cyber attacks on system performance. Specifically, this article proposes a hybrid control law containing multiple controllers, each of which is designed to counter different types of cyber attacks. By using the stochastic analysis approach, two sufficient criteria are provided to guarantee that the time-varying MASs satisfy the finite horizon H∞ consensus performance. Then, the controller parameters are obtained by solving the recursive linear matrix inequality. The usefulness of the theoretic results presented is demonstrated via a numerical example that contains a performance comparison of different secure control schemes.

This article investigates the resilient event-triggered (ET) distributed state estimation problem for nonlinear systems under denial-of-service (DoS) attacks. Different from the existing results mainly considering linear or specified nonlinear systems, more general nonlinear systems are considered in this study. Moreover, the considered DoS attacks are able to compromise different communication links among estimators independently. In this context, by resorting to the techniques of incremental homogeneity, a nonlinear ET distributed estimation scheme is designed to estimate the states and regulate the data transmission. Under this scheme, the resilient state estimation is achieved by employing a multimode switching estimator, and the problem of efficiency loss of the ET mechanism caused by DoS attacks is solved by designing a dynamic trigger threshold with switched update laws. Then, based on the decay rates of the Lyapunov function corresponding to different communication modes, sufficient conditions are given to guarantee the stability of the estimation error system under DoS attacks. Finally, simulation results are provided to verify the effectiveness of the proposed method.

Shor's quantum algorithm and other efficient quantum algorithms can break many public-key cryptographic schemes in polynomial time on a quantum computer. In response, researchers proposed postquantum cryptography to resist quantum computers. The multivariate cryptosystem (MVC) is one of a few options of postquantum cryptography. It is based on the NP-hardness of the computational problem to solve nonlinear equations over a finite field. Recently, Wang et al. (2018) proposed a MVC based on extended clipped hopfield neural networks (eCHNN). Its main security assumption is backed by the discrete logarithm (DL) problem over Matrics. In this brief, we present quantum cryptanalysis of Wang et al. 's eCHNN-based MVC. We first show that Shor's quantum algorithm can be modified to solve the DL problem over Matrics. Then we show that Wang et al. 's construction of eCHNN-based MVC is not secure against quantum computers; this against the original intention of that multivariate cryptography is one of a few options of postquantum cryptography.

This article studies the pinning synchronization problem with edge-based decentralized adaptive schemes under link attacks. The link attacks considered here are a class of malicious attacks to break links between neighboring nodes in complex networks. In such an insecure network environment, two kinds of edge-based decentralized adaptive update strategies (synchronous and asynchronous) on coupling strengths and gains are designed to realize the security synchronization of complex networks. Moreover, by virtue of the edge pinning technique, the corresponding secure synchronization problem is considered under the case where only a small fraction of coupling strengths and gains is updated. These designed adaptive strategies do not require any global information, and therefore, the obtained results in this article are developed in a fully decentralized framework. Finally, a numerical example is provided to verify the availability of the achieved theoretical outcomes.

The aggregate signature is a special signature mode widely used in the wireless medical sensor network. The aggregate signature can reduce bandwidth and storage space in this sensor network by aggregating multiple signatures from different signers into a short signature. Traditional aggregate signature protocols based on the discrete logarithm problem and the great integer factorization problem cannot resist quantum attacks. In this article, we propose a new noninteractive identity-based aggregate signature protocol from lattices. We prove that the proposed protocol is secure in the quantum random oracle model. Additionally, the performance of the proposed protocol is much better than most previous protocols according to our experiment. Meanwhile, the size of the aggregate signature in our protocol is a logarithmic function of the number of signatures being aggregated.

Sparse stealth attack constructions that minimize the mutual information between the state variables and the observations are proposed. The attack construction is formulated as the design of a multivariate Gaussian distribution that aims to minimize the mutual information while limiting the Kullback-Leibler divergence between the distribution of the observations under attack and the distribution of the observations without attack. The sparsity constraint is incorporated as a support constraint of the attack distribution. Two heuristic greedy algorithms for the attack construction are proposed. The first algorithm assumes that the attack vector consists of independent entries, and therefore, requires no communication between different attacked locations. The second algorithm considers correlation between the attack vector entries which results in better attack performance at the expense of coordination between different locations. We numerically evaluate the performance of the proposed attack constructions on IEEE test systems and show that it is feasible to construct stealth attacks that generate significant disruption with a low number of compromised sensors., Comment: 10 pages, 6 figures, submited to IEEE Trans. Smart Grid

Unconditionally secure quantum bit commitment (QBC) was widely believed to be impossible for more than two decades. But recently, based on an anomalous behavior found in quantum steering, we proposed a QBC protocol which can be unconditionally secure in principle. The protocol requires the use of infinite-dimensional systems, therefore it may seem less feasible in practice. Here we propose a quantum optical method based on Mach-Zehnder interferometer, which gives a very good approximation to such infinite-dimensional systems. Thus, it enables a proof-of-principle experimental implementation of our protocol, which can also serve as a practically secure QBC scheme. Other multi-party cryptographic protocols such as quantum coin tossing can be built upon it too. Our approach also reveals a relationship between infinity and non-locality, which may have an impact on the research of fundamental theories., Comment: Very close to the published version

With the development of IoT, more and more data is offloaded from the cloud to the edge for computing, eventually forming a collaborative computing model at the edge. However, in this model, the problem of secure data transmission has not been solved. In this model, data is transmitted and forwarded in multiple messaging systems, and existing security schemes cannot achieve end-to-end security in a multi-hop, broadcast transmission model. Therefore, in this paper, we propose a new security scheme based on proxy re-encryption and broadcast encryption techniques. Moreover, the performance and security of the scheme are further enhanced by using online-offline techniques and a trusted execution environment when integrating the scheme with edge collaboration. Finally, this paper proves the security of the scheme in theory, compares the functionality of the scheme, analyzes the theoretical performance of the scheme, and finally measures the actual performance of the scheme in the edge collaboration system.

A novel deterministic sensing matrix design approach applied to enable secured compressed sensing and transmission of fingerprint images. The performance of the sensing matrix was analyzed in detail by varying compressed sensing and security parameters. The number of sampling and sparse coefficient are the parameters taken under consideration from compressed sensing, whereas the encryption key is from the security scheme. The first stage in the performance study is acquiring the signal, and followed by sparsely modelling it using Orthogonal Matching Pursuit (OMP) algorithm. The sparse modelled data is multiplied by the proposed deterministic partial orthogonal Discrete Cosine Transform (DCT) sensing matrix to reduce its dimension and encrypt it. To introduce confusion on the DCT matrix rows, the pseudo-random permutation is applied to the DCT matrix rows before sensing matrix derivation. Additionally, recovering and decryption of the compressed image accomplished with the help of a convex optimization method. The results obtained from the simulation of the proposed system confirmed that a better Peak Signal to Noise Ratio (PSNR) than the recommended value for wireless transmission is archived using a sample below 25% without losing a significant number of fingerprint minutiae.

This article is concerned with the security-based resilient robust model predictive control (RMPC) problem for a class of discrete-time polytopic uncertain systems with deception attacks and the round-robin (RR) protocol. The well-known RR protocol is adopted with the hope to reduce the communication burden, where the control signal is transmitted only at the given transmission instant. By means of the concept of bounded energy and the independent Bernoulli-distributed (BD) white sequences, a representative attack model is given. A definition of mean-square (MS) security in H₂-sense is employed to reasonably explain the dynamics process of the controlled systems. We aim at designing a set of resilient RMPC controllers so as to make the closed-loop system state under consideration of deception attacks and the RR protocol enter and stay in a certain bounded region. Furthermore, with the aid of stochastic analysis methods and inequality analysis techniques, sufficient conditions are obtained to satisfy the desirable security requirements. To tackle the nonconvex obstacles resulting from attacks and the RR protocol, the cone complementary linearization (CCL) method is exploited to cast them into a convex optimization problem (OP) for its solvability. Then, an online OP regarding a certain upper bound of the concerned objective is provided for the solvability and resilient RMPC-based controller gains are obtained. Finally, two examples, including a high-purity distillation one and a numerical one, are used to demonstrate the effectiveness and the validity of the proposed techniques.

Traditional cryptographic approaches may not always be suitable for cyber-physical systems (CPSs). In this correspondence, we present a control-theoretic approach allowing a networked controller and a smart actuator of a stochastic CPS to agree on a common secret key without resorting to classical cryptographic approaches. More precisely, by utilizing the asymmetry in the system model knowledge available to the control system defender and to the eavesdropper, we propose a key-agreement scheme utilizing a simultaneous input and state estimation algorithm. The validity of the proposed solution is shown through a numerical example.

In this article, an efficient method for finite-time secure dynamic state estimation in cyber-physical systems (CPSs) subjected to unknown inputs and cyber-attacks is proposed. The proposed approach is based on a set of local finite-time state estimators operating over the subsets of sensory nodes, which are designed to estimate the states of the CPS subjected to unknown inputs. When cyber-attacks compromise some sensory nodes, the estimation results of the local finite-time estimators which use the measurements of the attacked sensors can be corrupted. An efficient detection algorithm is thus proposed to identify the valid local estimators that are completely devoid of the attacked sensory nodes. The information of the valid local estimators is then used to achieve secure state estimation and localization of the launched cyber-attack. The necessary and sufficient conditions for the feasibility and finite-time convergence of the proposed estimation mechanism are analytically derived and proven. The effectiveness of the proposed method is demonstrated by testing it on a dc electric motor. Likewise, some online software-in-the-loop tests are conducted to demonstrate the real-time feasibility of the proposed algorithm.

Deep learning-based auto-driving systems are vulnerable to adversarial examples attacks which may result in wrong decision making and accidents. An adversarial example can fool the well trained neural networks by adding barely imperceptible perturbations to clean data. In this paper, we explore the mechanism of adversarial examples and adversarial robustness from the perspective of statistical mechanics, and propose an statistical mechanics-based interpretation model of adversarial robustness. The state transition caused by adversarial training based on the theory of fluctuation dissipation disequilibrium in statistical mechanics is formally constructed. Besides, we fully study the adversarial example attacks and training process on system robustness, including the influence of different training processes on network robustness. Our work is helpful to understand and explain the adversarial examples problems and improve the robustness of deep learning-based auto-driving systems.

Image watermarking commonly involves singular value decomposition (SVD) because of its simplicity and minimal effect on image quality. However, SVD-based image watermarking schemes suffer from some drawbacks such as the false positive problem (FPP), and an undesirable trade-off between vital properties such as imperceptibility, embedding capacity, and robustness. To address these drawbacks, we improve upon the SVD-based color image watermarking by incorporating integer wavelet transform (IWT) and chaotic maps. A grayscale image watermark is decomposed into eight gray levels (bit-planes) before being encrypted and re-ordered by a chaotic sequence. Each encrypted bit-plane is inserted into singular values of the sub-bands of a host image. The embedding process is controlled by chaos-based multiple scaling factors (MSF) which are updated in each embedded bit-plane. The resulting values are involved in generating a hash value at the end of the embedding processes. The hash value is used to overcome FPP issues and improves security. Our findings illustrate the proposed scheme’s robustness, security, imperceptibility, and capacity. It also exhibits excellent robustness against attacks and its performance surpasses a variety of existing schemes. In addition, the proposed scheme is hypersensitive to even the slightest secret key variations.

A secure distributed estimation method against data integrity attacks is presented for general Internet-of-Things (IoT) systems. By exploiting the spatial sparsity of the attacks, a robust optimal estimation objective to protect the entire IoT system is established. Then, a false data processor is developed to suppress the bad effects of the attacks. The convergence of the presented method is analyzed. It shows that, under accessible conditions, the estimation error will converge to be uniformly bounded for all cases, i.e., all communication links and transmitted data are arbitrarily compromised. Furthermore, by utilizing the property of persistence, a quantitative trust model is established for IoT systems based on the past behaviors of the nodes. Then, a trust-based distributed estimation method is proposed to enhance security. Simulation results by considering an IoT system with 50 nodes and 145 edges are also provided to verify our theoretical results.

This article studies the resilient control problem for a class of unknown nonlinear systems with fading measurements under malicious denial-of-service (DoS) attacks. The system output is assumed to be transmitted through a fading channel, where the fading phenomenon is described by a Rice fading model. The strategy of the attacker is to periodically interfere with the networked channels to reduce the success rate of data transmissions. First, a dynamic linearization method along the iteration domain is introduced to convert the nonlinear system into an equivalent data-related model. Then, a model-free adaptive iterative learning control (MFAILC) scheme is presented, which is independent of model information. The convergence of the MFAILC scheme is deduced theoretically and the influence of DoS attacks and stochastic fading phenomenon on system stability are also analyzed. Finally, the effectiveness of the design is verified by a numerical simulation and a trajectory tracking example of wheeled mobile robots (WMRs).

We present a method to search for a probe (or query) image representation against a large gallery in the encrypted domain. We require that the probe and gallery images be represented in terms of a fixed-length representation, which is typical for representations obtained from learned networks. Our encryption scheme is agnostic to how the fixed-length representation is obtained and can therefore be applied to any fixed-length representation in any application domain. Our method, dubbed HERS (Homomorphically Encrypted Representation Search), operates by (i) compressing the representation towards its estimated intrinsic dimensionality with minimal loss of accuracy (ii) encrypting the compressed representation using the proposed fully homomorphic encryption scheme, and (iii) efficiently searching against a gallery of encrypted representations directly in the encrypted domain, without decrypting them. Numerical results on large galleries of face, fingerprint, and object datasets such as ImageNet show that, for the first time, accurate and fast image search within the encrypted domain is feasible at scale (500 seconds; $275\times$ speed up over state-of-the-art for encrypted search against a gallery of 100 million). Code is available at https://github.com/human-analysis/hers-encrypted-image-search, Comment: Published in the Trustworthy Biometrics Special Issue of IEEE Transactions on Biometrics, Behavior, and Identity Science 2021

The authentication and content protection of multimedia data is a challenging task in the present scenario. One solution is to generate the perceptual hash which essentially authenticates the multimedia data and can also be dealt with image database search problems. In this paper, a novel system for generating an image hash is presented. The proposed system utilizes the global and local features in the hash generation process. The local features are obtained from non-linear scale-space based KAZE features. These KAZE features have the ability to capture the most stable point under several content preserving distortions. In contrast, first the input image is converted to a normalized image, it is transformed to the log-polar coordinate system, and then a reference image using the local contrast in the wavelet domain is obtained to extrcat the global features. An intermediate hash sequence is achieved by extracting the significant information from the reference image using the singular value decomposition. The final hash sequence combines both the vectors followed by the randomization process. Extensive experiments are conducted to demonstrate the feasibility and the robustness of the proposed hashing system against a wide range of intentional/unintentional distortions. Further, the comparative analysis with some state-of-the-art techniques validates the better discrimination of the proposed work.

In this article, we study a variation of the active target-attacker-defender (ATAD) differential game involving multiple targets, an attacker, and a defender. Our model allows for 1) a capability of the defender to switch roles from rescuer (rendezvous with all the targets) to interceptor (intercepts the attacker) and vice versa and 2) the attacker to continuously pursue the closest target (which can change during the course of the game). We assume that the mode of the defender (rescue or interception) defines the mode of the game itself. Using the framework of Games of a Degree, we first analyze the game within each mode. More specifically, the objectives of the players are taken as a combination of weighted Euclidean distances and penalties on their control efforts. We model the interaction of the players within each mode as a linear quadratic differential game (LQDG) and obtain the open-loop Nash equilibrium strategies. We then use the receding horizon approach to enable switching between the modes to obtain switching strategies for the players. By partitioning the matrices associated with the Riccati differential equations we obtain geometric characterization of the trajectories of the players. Furthermore, under mild restrictions on the problem parameters and for a particular choice of the defender's switching function we show that interception mode is invariant. We illustrate our results with numerical simulations. Experimental results involving multiple autonomous differential drive mobile robots are presented.

While input or output-perturbation-based adversarial training techniques have been exploited to enhance the generalization capability of a variety of nonfuzzy and fuzzy classifiers by means of dynamic regularization, their performance may perhaps be very sensitive to some inappropriate adversarial samples. In order to avoid this weakness and simultaneously ensure enhanced generalization capability, this work attempts to explore a novel knowledge adversarial attack model for the zero-order Tagaki-Sugeno-Kang (TSK) fuzzy classifiers. The proposed model is motivated by exploiting the existence of special knowledge adversarial attacks from the perspective of the human-like thinking process when training an interpretable zero-order TSK fuzzy classifier. Without any direct use of adversarial samples, which is different from input or output perturbation-based adversarial attacks, the proposed model considers adversarial perturbations of interpretable zero-order fuzzy rules in a knowledge-oblivion and/or knowledge-bias or their ensemble to mimic the robust use of knowledge in the human thinking process. Through dynamic regularization, the proposed model is theoretically justified for its strong generalization capability. Accordingly, a novel knowledge adversarial training method called KAT is devised to achieve promising generalization performance, interpretability, and fast training for zero-order TSK fuzzy classifiers. The effectiveness of KAT is manifested by the experimental results on 15 benchmarking UCI and KEEL datasets.

In this paper, we investigate the behavior of agents with bounded rationality, attacking a set of stochastic sensors measuring the state of a binary event. The coordination problem between the attackers is formulated as a multi-player non-zero-sum one-shot game. Each attacker aims to maximize the probability that a certain detector will produce an erroneous estimate of the true event, while they remain stealthy. To better predict the outcome of this game, we categorize the players based on the number of strategic thinking steps they will take. Each level-k attacker behaves based on subjective beliefs of the others' behaviors, which are quantified via a Poisson distribution over the lower levels. The expected best responses for the type of each attacker are derived. The limiting conditions as the cognitive levels increase, as well as when the attackers fully coordinate, are shown to converge to the Nash equilibrium.

Connected vehicles are potential solutions to address some of the existing challenges in transportation systems, such as emission, traffic congestion, and fuel consumption. However, vehicular communication networks endure from reliability and security issues. Cyber-attacks with performance-disrupting purposes can lead to catastrophic collisions in connected vehicles and increase traffic congestion. To ensure safety and reliability, a monitoring system with the capability of diagnosing cyber-attacks is necessary for connected vehicles. In this study, we consider a platoon of connected vehicles equipped with cooperative adaptive cruise control (CACC) that is subjected to a specific type of cyber-attack, namely ``False Data Injection'' (FDI) attack. A smart FDI attack is modeled with ghost vehicles injection into the connected vehicles network to disrupt the performance of the whole system. To ease the analysis, we develop a partial differential equation (PDE) model for the CACC vehicle dynamics using its ordinary differential equations (ODE). Furthermore, a PDE observer is designed to detect the FDI attack and locate the attack's injection point in the platoon. Lyapunov stability theory has been utilized to verify the observer's convergence under no-attack scenario and study residuals' behaviors in the presence of the attack. A non-zero constant threshold is considered to provide robustness in attack detection despite the measurement noises. Eventually, the effectiveness of the proposed algorithm is evaluated with simulation studies.

Active Directory is the default security management system for Windows domain networks. We study the shortest path edge interdiction problem for defending Active Directory style attack graphs. The problem is formulated as a Stackelberg game between one defender and one attacker. The attack graph contains one destination node and multiple entry nodes. The attacker's entry node is chosen by nature. The defender chooses to block a set of edges limited by his budget. The attacker then picks the shortest unblocked attack path. The defender aims to maximize the expected shortest path length for the attacker, where the expectation is taken over entry nodes. We observe that practical Active Directory attack graphs have small maximum attack path length and are structurally close to trees. We first show that even if the maximum attack path length is a constant, the problem is still w[1]-hard with respect to the defender's budget. Having a small maximum attack path length and a small budget is not enough to design fixed-parameter algorithms. If we further assume that the number of entry nodes is small, then we derive a fixed-parameter tractable algorithm. We then propose two other fixed-parameter algorithms by exploiting the tree-like features. One is based on tree decomposition and requires a small tree width. The other assumes a small number of splitting nodes (nodes with multiple out-going edges). Finally, the last algorithm is converted into a graph convolutional neural network based heuristic, which scales to larger graphs with more splitting nodes.