Biblio

CAPTCHAs are designed to prevent malicious bot programs from abusing websites. Most online service providers deploy audio CAPTCHAs as an alternative to text and image CAPTCHAs for visually impaired users. However, prior research investigating the security of audio CAPTCHAs found them highly vulnerable to automated attacks using Automatic Speech Recognition (ASR) systems. To improve the robustness of audio CAPTCHAs against automated abuses, we present the design and implementation of an audio adversarial CAPTCHA (aaeCAPTCHA) system in this paper. The aaeCAPTCHA system exploits audio adversarial examples as CAPTCHAs to prevent the ASR systems from automatically solving them. Furthermore, we conducted a rigorous security evaluation of our new audio CAPTCHA design against five state-of-the-art DNN-based ASR systems and three commercial Speech-to-Text (STT) services. Our experimental evaluations demonstrate that aaeCAPTCHA is highly secure against these speech recognition technologies, even when the attacker has complete knowledge of the current attacks against audio adversarial examples. We also conducted a usability evaluation of the proof-of-concept implementation of the aaeCAPTCHA scheme. Our results show that it achieves high robustness at a moderate usability cost compared to normal audio CAPTCHAs. Finally, our extensive analysis highlights that aaeCAPTCHA can significantly enhance the security and robustness of traditional audio CAPTCHA systems while maintaining similar usability.

Our position is that a key component of securing cyber-physical systems (CPS) is to develop a theory of accountability that encompasses both control and computing systems. We envision that a unified theory of accountability in CPS can be built on a foundation of causal information flow analysis. This theory will support design and analysis of mechanisms at various stages of the accountability regime: attack detection, responsibility-assignment (e.g., attack identification or localization), and corrective measures (e.g., via resilient control) As an initial step in this direction, we summarize our results on attack detection in control systems. We use the Kullback-Liebler (KL) divergence as a causal information flow measure. We then recover, using information flow analyses, a set of existing results in the literature that were previously proved using different techniques. These results cover passive detection, stealthy attack characterization, and active detection. This research direction is related to recent work on accountability in computational systems [1], [2], [3], [4]. We envision that by casting accountability theories in computing and control systems in terms of causal information flow, we can provide a common foundation to develop a theory for CPS that compose elements from both domains.

A novel deep neural network is proposed, for accurate and robust crowd counting. Crowd counting is a complex task, as it strongly depends on the deployed camera characteristics and, above all, the scene perspective. Crowd counting is essential in security applications where Internet of Things (IoT) cameras are deployed to help with crowd management tasks. The complexity of a scene varies greatly, and a medium to large scale security system based on IoT cameras must cater for changes in perspective and how people appear from different vantage points. To address this, our deep architecture extracts multi-scale features with a pyramid contextual module to provide long-range contextual information and enlarge the receptive field. Experiments were run on three major crowd counting datasets, to test our proposed method. Results demonstrate our method supersedes the performance of state-of-the-art methods.

Interconnect opens are known to be one of the predominant defects in nanoscale technologies. Automatic test pattern generation for open faults is challenging, because of their rather unstable behavior and the numerous electrical parameters which need to be considered. Thus, most approaches try to avoid accurate modeling of all constraints like the influence of the aggressors on the open net and use simplified fault models in order to detect as many faults as possible or make assumptions which decrease both complexity and accuracy. Yet, this leads to the problem that not only generated tests may be invalidated but also the localization of a specific fault may fail - in case such a model is used as basis for diagnosis. Furthermore, most of the models do not consider the problem of oscillating behavior, caused by feedback introduced by coupling capacitances, which occurs in almost all designs. In [1], the Robust Enhanced Aggressor Victim Model (REAV) and in [2] an extension to address the problem of oscillating behavior were introduced. The resulting model does not only consider the influence of all aggressors accurately but also guarantees robustness against oscillating behavior as well as process variations affecting the thresholds of gates driven by an open interconnect. In this work we present the first diagnostic classification algorithm for this model. This algorithm considers all constraints enforced by the REAV model accurately - and hence handles unknown values as well as oscillating behavior. In addition, it allows to distinguish faults at the same interconnect and thus reducing the area that has to be considered for physical failure analysis. Experimental results show the high efficiency of the new method handling circuits with up to 500,000 non-equivalent faults and considerably increasing the diagnostic resolution.

We introduce Active Dependency Mapping (ADM), a method for establishing dependency relations among a set of interdependent services. The approach is to artificially degrade network performance to infer which assets on the network support a particular process. Artificial degradation of the network environment could be transparent to users; run continuously it could identify dependencies that are rare or occur only at certain timescales. A useful byproduct of this dependency analysis is a quantitative assessment of the resilience and robustness of the system. This technique is intriguing for hardening both enterprise networks and cyber physical systems. We present a proof-of-concept experiment executed on a real-world set of interrelated software services. We assess the efficacy of the approach, discuss current limitations, and suggest options for future development of ADM.

The Named-Data Networking (NDN) has emerged as a clean-slate Internet proposal on the wave of Information-Centric Networking. Although the NDN's data-plane seems to offer many advantages, e.g., native support for multicast communications and flow balance, it also makes the network infrastructure vulnerable to a specific DDoS attack, the Interest Flooding Attack (IFA). In IFAs, a botnet issuing unsatisfiable content requests can be set up effortlessly to exhaust routers' resources and cause a severe performance drop to legitimate users. So far several countermeasures have addressed this security threat, however, their efficacy was proved by means of simplistic assumptions on the attack model. Therefore, we propose a more complete attack model and design an advanced IFA. We show the efficiency of our novel attack scheme by extensively assessing some of the state-of-the-art countermeasures. Further, we release the software to perform this attack as open source tool to help design future more robust defense mechanisms.

With the remarkable success of deep learning, Deep Neural Networks (DNNs) have been applied as dominant tools to various machine learning domains. Despite this success, however, it has been found that DNNs are surprisingly vulnerable to malicious attacks; adding a small, perceptually indistinguishable perturbations to the data can easily degrade classification performance. Adversarial training is an effective defense strategy to train a robust classifier. In this work, we propose to utilize the generator to learn how to create adversarial examples. Unlike the existing approaches that create a one-shot perturbation by a deterministic generator, we propose a recursive and stochastic generator that produces much stronger and diverse perturbations that comprehensively reveal the vulnerability of the target classifier. Our experiment results on MNIST and CIFAR-10 datasets show that the classifier adversarially trained with our method yields more robust performance over various white-box and black-box attacks.

Malicious software, known as malware, has become urgently serious threat for computer security, so automatic mal-ware classification techniques have received increasing attention. In recent years, deep learning (DL) techniques for computer vision have been successfully applied for malware classification by visualizing malware files and then using DL to classify visualized images. Although DL-based classification systems have been proven to be much more accurate than conventional ones, these systems have been shown to be vulnerable to adversarial attacks. However, there has been little research to consider the danger of adversarial attacks to visualized image-based malware classification systems. This paper proposes an adversarial attack method based on the gradient to attack image-based malware classification systems by introducing perturbations on resource section of PE files. The experimental results on the Malimg dataset show that by a small interference, the proposed method can achieve success attack rate when challenging convolutional neural network malware classifiers.

Modulation classification is an important component of cognitive self-driving networks. Recently many ML-based modulation classification methods have been proposed. We have evaluated the robustness of 9 ML-based modulation classifiers against the powerful Carlini & Wagner (C-W) attack and showed that the current ML-based modulation classifiers do not provide any deterrence against adversarial ML examples. To the best of our knowledge, we are the first to report the results of the application of the C-W attack for creating adversarial examples against various ML models for modulation classification.

Robustness, both to accident and to malevolent perturbations, is a crucial determinant of the successful deployment of deep convolutional neural network based SAR ATR systems in various security-sensitive applications. This paper performs a detailed adversarial robustness evaluation of deep convolutional neural network based SAR ATR models across two public available SAR target recognition datasets. For each model, seven different adversarial perturbations, ranging from gradient based optimization to self-supervised feature distortion, are generated for each testing image. Besides adversarial average recognition accuracy, feature attribution techniques have also been adopted to analyze the feature diffusion effect of adversarial attacks, which promotes the understanding of vulnerability of deep learning models.

Recent advances in natural language generation have introduced powerful language models with high-quality output text. However, this raises concerns about the potential misuse of such models for malicious purposes. In this paper, we study natural language watermarking as a defense to help better mark and trace the provenance of text. We introduce the Adversarial Watermarking Transformer (AWT) with a jointly trained encoder-decoder and adversarial training that, given an input text and a binary message, generates an output text that is unobtrusively encoded with the given message. We further study different training and inference strategies to achieve minimal changes to the semantics and correctness of the input text.AWT is the first end-to-end model to hide data in text by automatically learning -without ground truth- word substitutions along with their locations in order to encode the message. We empirically show that our model is effective in largely preserving text utility and decoding the watermark while hiding its presence against adversaries. Additionally, we demonstrate that our method is robust against a range of attacks.

Secret-related carrier intelligent management and control system uses the Internet of Things and artificial intelligence to solve the transformation of secret-related carrier management and control from manual operation to automatic detection, precise monitoring, and intelligent decision-making, and use technical means to resolve security risks. However, the coexistence of multiple heterogeneous networks will lead to various network security problems in the secret carrier intelligent management and control. Aiming at the actual requirements of the intelligent management and control of secret-related carriers, this paper proposes a system structure including device domain, network domain, platform domain and user domain, and conducts a detailed system security analysis, and introduces intelligent agent technology, and proposes a distributed system. The hierarchical system structure of the secret-related carrier intelligent management and control security protection model has good robustness and portability.

In this short paper, we survey some work at the intersection of Artificial Intelligence (AI) and security that are based on game theoretic considerations, and particularly focus on the author's (our) contribution in these areas. One half of this paper focuses on applications of game theoretic and learning reasoning for addressing security applications such as in public safety and wildlife conservation. In the second half, we present recent work that attacks the learning components of these works, leading to sub-optimal defense allocation. We finally end by pointing to issues and potential research problems that can arise due to data quality in the real world.

Despite the performance of state-of-the-art Artificial Intelligence (AI) systems, some sectors hesitate to adopt AI because of a lack of trust in these systems. This attitude is prevalent among high-risk areas, where there is a reluctance to remove humans entirely from the loop. In these scenarios, Augmentation provides a preferred alternative over complete Automation. Instead of replacing humans, AI Augmentation uses AI to improve and support human operations, creating an environment where humans work side by side with AI systems. In this paper, we discuss how AI Augmentation can provide a path for building Trustworthy AI. We exemplify this approach using Robot Teleoperation. We lay out design guidelines and motivations for the development of AI Augmentation for Robot Teleoperation. Finally, we discuss the design of a Robot Teleoperation testbed for the development of AI Augmentation systems.

In recent years, ICN (Information-Centric Networking) has been under the spotlight as a network that mainly focuses on transmitted and received data rather than on the hosts that transmit and receive data. Generally, the communication networks such as ICNs are required to be robust against network failures caused by attacks and disasters. One of the metrics for the robustness of conventional host-centric networks, e.g., TCP/IP network, is reachability between nodes in the network after network failures, whereas the key metric for the robustness of ICNs is content availability. In this paper, we focus on an arbitrary ICN network and derive the content availability for a given probability of node removal. Especially, we analytically obtain the average content availability over an entire network in the case where just a single path from a node to a repository, i.e., contents server, storing contents is available and where multiple paths to the repository are available, respectively. Furthermore, through several numerical evaluations, we investigate the effect of the structure of network topology as well as the pattern and scale of the network failures on the content availability in ICN. Our findings include that, regardless of patterns of network failures, the content availability is significantly improved by caching contents at routers and using multiple paths, and that the content availability is more degraded at cluster-based node removal compared with random node removal.

In this era, with a great extent of automation and connection, modern production processes are highly prone to cyber-attacks. The sensor-controller chain becomes an obvious target for attacks because sensors are commonly used to regulate production facilities. In this research, we introduce a new control configuration for the system, which is sensitive to time delay attacks (TDA), in which data transfer from the sensor to the controller is intentionally delayed. The attackers want to disrupt and damage the system by forcing controllers to use obsolete data about the system status. In order to improve the accuracy of delay identification and prediction, as well as erroneous limit and estimation for control, a new control structure is developed by an Internal Model Control (IMC) based Proportional-Integral-Derivative (PID) scheme with a fractional filter. An additional concept is included to mitigate the effect of time delay attack, i.e., the smith predictor. Simulation studies of the established control framework have been implemented with two numerical examples. The performance assessment of the proposed method has been done based on integral square error (ISE), integral absolute error (IAE) and total variation (TV).

Industrial Control Systems (ICS) are increasingly facing the threat of False Data Injection (FDI) attacks. As an emerging intrusion detection scheme for ICS, process-based Intrusion Detection Systems (IDS) can effectively detect the anomalies caused by FDI attacks. Specifically, such IDS establishes anomaly detection model which can describe the normal pattern of industrial processes, then perform real-time anomaly detection on industrial process data. However, this method suffers low detection accuracy due to the complexity and instability of industrial processes. That is, the process data inherently contains sophisticated nonlinear spatial-temporal correlations which are hard to be explicitly described by anomaly detection model. In addition, the noise and disturbance in process data prevent the IDS from distinguishing the real anomaly events. In this paper, we propose an Anomaly Detection approach based on Robust Spatial-temporal Modeling (AD-RoSM). Concretely, to explicitly describe the spatial-temporal correlations within the process data, a neural based state estimation model is proposed by utilizing 1D CNN for temporal modeling and multi-head self attention mechanism for spatial modeling. To perform robust anomaly detection in the presence of noise and disturbance, a composite anomaly discrimination model is designed so that the outputs of the state estimation model can be analyzed with a combination of threshold strategy and entropy-based strategy. We conducted extensive experiments on two benchmark ICS security datasets to demonstrate the effectiveness of our approach.

In order to realize Intelligent Disaster Recovery and break the traditional reactive backup mode, it is necessary to forecast the potential system anomalies, and proactively backup the real-time datas and configurations. System logs record the running status as well as the critical events (including errors and warnings), which can help to detect system performance, debug system faults and analyze the causes of anomalies. What's more, with the features of real-time, hierarchies and easy-access, log data can be an ideal source for monitoring system status. To reduce the complexity and improve the robustness and practicability of existing log-based anomaly detection methods, we propose a new anomaly detection mechanism based on hierarchical weights, which can deal with unstable log data. We firstly extract semantic information of log strings, and get the word-level weights by SIF algorithm to embed log strings into vectors, which are then feed into attention-based Long Short-Term Memory(LSTM) deep learning network model. In addition to get sentence-level weight which can be used to explore the interdependence between different log sequences and improve the accuracy, we utilize attention weights to help with building workflow to diagnose the abnormal points in the execution of a specific task. Our experimental results show that the hierarchical weights mechanism can effectively improve accuracy of perdition task and reduce complexity of the model, which provides the feasibility foundation support for Intelligent Disaster Recovery.

The popularity of cryptocurrencies has garnered interest from cybercriminals, spurring an onslaught of cryptojacking campaigns that aim to hijack computational resources for the purpose of mining cryptocurrencies. In this paper, we present a cross-stack cryptojacking defense system that spans the hardware and OS layers. Unlike prior work that is confined to detecting cryptojacking behavior within web browsers, our solution is application agnostic. We show that tracking instructions that are frequently used in cryptographic hash functions serve as reliable signatures for fingerprinting cryptojacking activity. We demonstrate that our solution is resilient to multi-threaded and throttling evasion techniques that are commonly employed by cryptojacking malware. We characterize the robustness of our solution by extensively testing a diverse set of workloads that include real consumer applications. Finally, an evaluation of our proof-of-concept implementation shows minimal performance impact while running a mix of benchmark applications.

Currently, the rapid development of digital communication and multimedia has made security an increasingly prominent issue of communicating, storing, and transmitting digital data such as images, audio, and video. Encryption techniques such as chaotic map based encryption can ensure high levels of security of data and have been used in many fields including medical science, military, and geographic satellite imagery. As a result, ensuring image data confidentiality, integrity, security, privacy, and authenticity while transferring and storing images over an unsecured network like the internet has become a high concern. There have been many encryption technologies proposed in recent years. This paper begins with a summary of cryptography and image encryption basics, followed by a discussion of different kinds of chaotic image encryption techniques and a literature review for each form of encryption. Finally, by examining the behaviour of numerous existing chaotic based image encryption algorithms, this paper hopes to build new chaotic based image encryption strategies in the future.

The software supply chain is a source of cybersecurity risk for many commercial and government organizations. Public data may be used to inform automated tools for detecting software supply chain risk during continuous integration and deployment. We link data from the National Vulnerability Database (NVD) with open version control data for the open source project OpenSSL, a widely used secure networking library that made the news when a significant vulnerability, Heartbleed, was discovered in 2014. We apply the Alhazmi-Malaiya Logistic (AML) model for software vulnerability discovery to this case. This model predicts a sigmoid cumulative vulnerability discovery function over time. Some versions of OpenSSL do not conform to the predictions of the model because they contain a temporary plateau in the cumulative vulnerability discovery plot. This temporary plateau feature is an empirical signature of a security failure mode that may be useful in future studies of software supply chain risk.

Blockchain technology (BCT) has paved a way for new potentials of handling serious data privacy, integrity and security issues in healthcare. To curb the increasing challenges in healthcare industry, healthcare organizations need to apply blockchain technology to better improve patient safety and protect patients records from counterfeiting and fraud. The purpose of this research paper was to define BCT can assist in improving pharmaceutical industries in Saudi Arabia upon utilization of its application. This study adopted quantitative methods to gather the study data. Based on healthcare leaders perception and Internet connection, lack of cooperation, and economic inequality were found to be leading factors hindering the application of blockchain technology in the pharmaceutical industries, Saudi Arabia. Factors facilitating the application of blockchain technology in the pharmaceutical industries, Saudi Arabia were found as system robustness of BCT, increased data safety and decentralization, need for enhanced supply chain management and interoperability, and government laws and policies. Adopting interventions that are targeted to specific patient population medications, effective delivery systems, transit provider reimbursement far from intensity and volume of services towards value and quality was found to compromise the pre-existent challenges and real capacity in healthcare system. Although the relationship between implementation of blockchain technology and cost spending is negative in the short-term, in the long run, the relationship is positive Blockchain helps in managing multiple levels in a more secure way, reduces paper work and amplifies verification inefficiency.

Suppose that you are at a music festival checking on an artist, and you would like to quickly know about the song that is being played (e.g., title, lyrics, album, etc.). If you have a smartphone, you could record a sample of the live performance and compare it against a database of existing recordings from the artist. Services such as Shazam or SoundHound will not work here, as this is not the typical framework for audio fingerprinting or query-by-humming systems, as a live performance is neither identical to its studio version (e.g., variations in instrumentation, key, tempo, etc.) nor it is a hummed or sung melody. We propose an audio fingerprinting system that can deal with live version identification by using image processing techniques. Compact fingerprints are derived using a log-frequency spectrogram and an adaptive thresholding method, and template matching is performed using the Hamming similarity and the Hough Transform.

This paper proposes a high-performance audio fingerprint extraction method for identifying TV commercial advertisement. In the proposed method, a salient audio peak pair fingerprints based on constant Q transform (CQT) are hashed and stored, to be efficiently compared to one another. Experimental results confirm that the proposed method is quite robust in different noise conditions and improves the accuracy of the audio fingerprinting system in real noisy environments.

Ransomware, a class of self-propagating malware that uses encryption to hold the victims' data ransom, has emerged in recent years as one of the most dangerous cyber threats, with widespread damage; e.g., zero-day ransomware WannaCry has caused world-wide catastrophe, from knocking U.K. National Health Service hospitals offline to shutting down a Honda Motor Company in Japan [1]. Our close collaboration with security operations of large enterprises reveals that defense against ransomware relies on tedious analysis from high-volume systems logs of the first few infections. Sandbox analysis of freshly captured malware is also commonplace in operation. We introduce a method to identify and rank the most discriminating ransomware features from a set of ambient (non-attack) system logs and at least one log stream containing both ambient and ransomware behavior. These ranked features reveal a set of malware actions that are produced automatically from system logs, and can help automate tedious manual analysis. We test our approach using WannaCry and two polymorphic samples by producing logs with Cuckoo Sandbox during both ambient, and ambient plus ransomware executions. Our goal is to extract the features of the malware from the logs with only knowledge that malware was present. We compare outputs with a detailed analysis of WannaCry allowing validation of the algorithm's feature extraction and provide analysis of the method's robustness to variations of input data—changing quality/quantity of ambient data and testing polymorphic ransomware. Most notably, our patterns are accurate and unwavering when generated from polymorphic WannaCry copies, on which 63 (of 63 tested) antivirus (AV) products fail.