Biblio

Recent research has shown that machine learning models are susceptible to adversarial examples, allowing attackers to trick a machine learning model into making a mistake and producing an incorrect output. Adversarial examples are commonly constructed or discovered by using gradient-based methods that require white-box access to the model. In most real-world AI system deployments, having complete access to the machine learning model is an unrealistic threat model. However, it is possible for an attacker to construct adversarial examples even in the black-box case - where we assume solely a query capability to the model - with a variety of approaches each with its advantages and shortcomings. We introduce AutoAttacker, a novel reinforcement learning framework where agents learn how to operate around the black-box model by querying it, to effectively extract the underlying decision behaviour, and to undermine it successfully. AutoAttacker is a first of kind framework that uses reinforcement learning and assumes nothing about the differentiability or structure of the underlying function and is thus robust towards common defenses like gradient obfuscation or adversarial training. Finally, without differentiable output, as in binary classification, most methods cease to operate and require either an approximation of the gradient, or another approach altogether. Our approach, however, maintains the capability to function when the output descriptiveness diminishes.

Physical-layer security (PLS) has raised the attention of the research community in recent years, particularly for Internet of things (IoT) applications. Despite the use of classical cryptography, PLS provides security at physical layer, regardless of the computational power owned by the attacker. The investigations on PLS are numerous in the literature, but one main issue seems to be kept apart: how to measure the benefit that PLS can bring to cryptography? This paper tries to answer this question with an initial performance analysis of PLS in conjunction with typical cryptography of wireless communication protocols. Our results indicate that PLS can help cryptography to harden the attacker job in real operative scenario: PLS can increase the detection errors at the attacker's receiver, leading to inability to recover the cipher key, even if the plaintext is known.

The technological evolution brought by the Internet of things (IoT) comes with new forms of cyber-attacks exploiting the complexity and heterogeneity of IoT networks, as well as, the existence of many vulnerabilities in IoT devices. The detection of compromised devices, as well as the collection and preservation of evidence regarding alleged malicious behavior in IoT networks, emerge as areas of high priority. This paper presents a blockchain-based solution, which is designed for the smart home domain, dealing with the collection and preservation of digital forensic evidence. The system utilizes a private forensic evidence database, where the captured evidence is stored, along with a permissioned blockchain that allows providing security services like integrity, authentication, and non-repudiation, so that the evidence can be used in a court of law. The blockchain stores evidences' metadata, which are critical for providing the aforementioned services, and interacts via smart contracts with the different entities involved in an investigation process, including Internet service providers, law enforcement agencies and prosecutors. A high-level architecture of the blockchain-based solution is presented that allows tackling the unique challenges posed by the need for digitally handling forensic evidence collected from IoT networks.

The Internet, the de facto platform for large-scale content distribution, suffers from two issues that limit its manageability, efficiency, and evolution. First, the IP-based Internet is host-centric and agnostic to the content being delivered and, second, the tight coupling of the control and data planes restrict its manageability, and subsequently the possibility to create dynamic alternative paths for efficient content delivery. Here, we present the CURLING system that leverages the emerging Information-Centric Networking paradigm for enabling cost-efficient Internet-scale content delivery by exploiting multicasting and in-network caching. Following the software-defined networking concept that decouples the control and data planes, CURLING adopts an interdomain hop-by-hop content resolution mechanism that allows network operators to dynamically enforce/change their network policies in locating content sources and optimizing content delivery paths. Content publishers and consumers may also control content access according to their preferences. Based on both analytical modeling and simulations using real domain-level Internet subtopologies, we demonstrate how CURLING supports efficient Internet-scale content delivery without the necessity for radical changes to the current Internet.

Generative Adversarial Networks (GANs) are one of the well-known models to generate synthetic data including images, especially for research communities that cannot use original sensitive datasets because they are not publicly accessible. One of the main challenges in this area is to preserve the privacy of individuals who participate in the training of the GAN models. To address this challenge, we introduce a Differentially Private Conditional GAN (DP-CGAN) training framework based on a new clipping and perturbation strategy, which improves the performance of the model while preserving privacy of the training dataset. DP-CGAN generates both synthetic data and corresponding labels and leverages the recently introduced Renyi differential privacy accountant to track the spent privacy budget. The experimental results show that DP-CGAN can generate visually and empirically promising results on the MNIST dataset with a single-digit epsilon parameter in differential privacy.

Inductive contactless energy transfer (CET) systems show a certain oscillating transient behavior of inrush currents on both system sides. This causes current overshoots in the electrical components and has to be considered for the system dimensioning. This paper presents a simple and yet very accurate model, which describes the dynamic behavior of series-series compensated inductive CET systems. This model precisely qualifies the systems current courses for both sides in time domain. Additionally, an analysis in frequency domain allows further knowledge for parameter estimation. Since this model is applicable for purely resistive loads and constant voltage loads with bridge rectifiers, it is very practicable and can be useful for control techniques and narameter estimation.

The Internet of Things (IoT) is a technological vision in which constrained or embedded devices connect together through the Internet. This enables common objects to be empowered with communication and cooperation capabilities. Industry can take an enormous advantage of IoT, leading to the so-called Industrial IoT. In these systems, integrity, confidentiality, and access control over data are key requirements. An emerging approach to reach confidentiality and access control is Attribute-Based Encryption (ABE), which is a technique able to enforce cryptographically an access control over data. In this paper, we propose fABElous, an ABE scheme suitable for Industrial IoT applications which aims at minimizing the overhead of encryption on communication. fABElous ensures data integrity, confidentiality, and access control, while reducing the communication overhead of 35% with respect to using ABE techniques naively.

In this paper, we present an extensive evaluation of face recognition and verification approaches performed by the European COST Action MULTI-modal Imaging of FOREnsic SciEnce Evidence (MULTI-FORESEE). The aim of the study is to evaluate various face recognition and verification methods, ranging from methods based on facial landmarks to state-of-the-art off-the-shelf pre-trained Convolutional Neural Networks (CNN), as well as CNN models directly trained for the task at hand. To fulfill this objective, we carefully designed and implemented a realistic data acquisition process, that corresponds to a typical face verification setup, and collected a challenging dataset to evaluate the real world performance of the aforementioned methods. Apart from verifying the effectiveness of deep learning approaches in a specific scenario, several important limitations are identified and discussed through the paper, providing valuable insight for future research directions in the field.

The Internet of Things (IoT) is enabling a new generation of innovative services based on the seamless integration of smart objects into information systems. Such IoT devices generate an uninterrupted flow of information that can be transmitted through an untrusted network and stored on an untrusted infrastructure. The latter raises new security and privacy challenges that require novel cryptographic methods. Attribute-Based Encryption (ABE) is a new type of public-key encryption that enforces a fine-grained access control on encrypted data based on flexible access policies. The feasibility of ABE adoption in fully-fledged computing systems, i.e. smartphones or embedded systems, has been demonstrated in recent works. In this paper we assess the feasibility of the adoption of ABE in typical IoT constrained devices, characterized by limited capabilities in terms of computing, storage and power. Specifically, an implementation of three ABE schemes for ESP32, a low-cost popular platform to deploy IoT devices, is developed and evaluated in terms of encryption/decryption time and energy consumption. The performance evaluation shows that the adoption of ABE on constrained devices is feasible, although it has a cost that increases with the number of attributes. The analysis in particular highlights how ABE has a significant impact in the lifetime of battery-powered devices, which is impaired significantly when a high number of attributes is adopted.

The paper describes the process of coding information in the heterogenic multi-core system on chip for virtual-based security designed For image processing, signal processing and neural networks emulation. The coding of information carried out in assembly language according to the GOST. This is an implementation of the GOST - a standard symmetric key block cipher has a 64-bit block size and 256-bit key size.

Data privacy is a very important problem to address while sharing data among multiple organizations and has become very crucial in the health sectors since multiple organizations such as hospitals are storing data of patients in the form of Electronic Health Records. Stored data is used with other organizations or research analysts to improve the health care of patients. However, the data records contain sensitive information such as age, sex, and date of birth of the patients. Revealing sensitive data can cause a privacy breach of the individuals. This has triggered research that has led to many different privacy preserving techniques being introduced. Thus, we designed a technique that not only encrypts / hides the sensitive information but also sends the data to different organizations securely. To encrypt sensitive data we use different fuzzy logic membership functions. We then use an autoencoder neural network to send the modified data. The output data of the autoencoder can then be used by different organizations for research analysis.

Back-end-of-line compatible integration of NdFeB-based micro magnets onto 8 inch Si substrates is presented. Substrate conditioning procedures to enable further processing in a cleanroom environment are discussed. It is shown that permanent magnetic structures with lateral dimensions between 25μm and 2000μm and a depth up to 500μm can be fabricated reliably and reproducibly with a remanent magnetization of 340mT at a standard deviation as low as 5% over the substrate. To illustrate post-processing capabilities, the fabrication of micro magnet arrangements embedded in silicon frames is described.

Good programming languages provide helpful abstractions for writing secure code, but the security properties of the source language are generally not preserved when compiling a program and linking it with adversarial code in a low-level target language (e.g., a library or a legacy application). Linked target code that is compromised or malicious may, for instance, read and write the compiled program's data and code, jump to arbitrary memory locations, or smash the stack, blatantly violating any source-level abstraction. By contrast, a fully abstract compilation chain protects source-level abstractions all the way down, ensuring that linked adversarial target code cannot observe more about the compiled program than what some linked source code could about the source program. However, while research in this area has so far focused on preserving observational equivalence, as needed for achieving full abstraction, there is a much larger space of security properties one can choose to preserve against linked adversarial code. And the precise class of security properties one chooses crucially impacts not only the supported security goals and the strength of the attacker model, but also the kind of protections a secure compilation chain has to introduce. We are the first to thoroughly explore a large space of formal secure compilation criteria based on robust property preservation, i.e., the preservation of properties satisfied against arbitrary adversarial contexts. We study robustly preserving various classes of trace properties such as safety, of hyperproperties such as noninterference, and of relational hyperproperties such as trace equivalence. This leads to many new secure compilation criteria, some of which are easier to practically achieve and prove than full abstraction, and some of which provide strictly stronger security guarantees. For each of the studied criteria we propose an equivalent “property-free” characterization that clarifies which proof techniques apply. For relational properties and hyperproperties, which relate the behaviors of multiple programs, our formal definitions of the property classes themselves are novel. We order our criteria by their relative strength and show several collapses and separation results. Finally, we adapt existing proof techniques to show that even the strongest of our secure compilation criteria, the robust preservation of all relational hyperproperties, is achievable for a simple translation from a statically typed to a dynamically typed language.

This work presents the design and implementation of a large curved display system in a virtual reality (VR) environment that supports visualization of 2D datasets (e.g., images, buttons and text). By using this system, users are allowed to interact with data in front of a wide field of view and gain a high level of perceived immersion. We exhibit two use cases of this system, including (1) a virtual image wall as the display component of a 3D user interface, and (2) an inventory interface for a VR-based educational game. The use cases demonstrate capability and flexibility of curved displays in supporting varied purposes of data interaction within virtual environments.

The advancement and penetration of distributed energy resources (DERs) and renewable energy sources (RES) are transforming legacy energy systems in an attempt to reduce carbon emissions and energy waste. Demand Response (DR) has been identified as a key enabler of integrating these, and other, Smart Grid technologies, while, simultaneously, ensuring grid stability and secure energy supply. The massive deployment of smart meters, IoT devices and DERs dictate the need to move to decentralized, or even localized, DR schemes in the face of the increased scale and complexity of monitoring and coordinating the actors and devices in modern smart grids. Furthermore, there is an inherent need to guarantee interoperability, due to the vast number of, e.g., hardware and software stakeholders, and, more importantly, promote trust and incentivize the participation of customers in DR schemes, if they are to be successfully deployed.In this work, we illustrate the design of an energy system that addresses all of the roadblocks that hinder the large scale deployment of DR services. Our DR framework incorporates modern Smart Grid technologies, such as fog-enabled and IoT devices, DERs and RES to, among others, automate asset handling and various time-consuming workflows. To guarantee interoperability, our system employs OpenADR, which standardizes the communication of DR signals among energy stakeholders. Our approach acknowledges the need for decentralization and employs blockchains and smart contracts to deliver a secure, privacy-preserving, tamper-resistant, auditable and reliable DR framework. Blockchains provide the infrastructure to design innovative DR schemes and incentivize active consumer participation as their aforementioned properties promote transparency and trust. In addition, we harness the power of smart contracts which allows us to design and implement fully automated contractual agreements both among involved stakeholders, as well as on a machine-to-machine basis. Smart contracts are digital agents that "live" in the blockchain and can encode, execute and enforce arbitrary agreements. To illustrate the potential and effectiveness of our smart contract-based DR framework, we present a case study that describes the exchange of DR signals and the autonomous instantiation of smart contracts among involved participants to mediate and monitor transactions, enforce contractual clauses, regulate energy supply and handle payments/penalties.

The growth in Internet usage has increased the use of electronic services requiring users to register their identity on each service they subscribe to. This has resulted in the prevalence of redundant users data on different services. To protect and regulate access by users to these services identity management systems (IdMs)are put in place. IdMs uses frameworks and standards e.g SAML, OAuth and Shibboleth to manage digital identities of users for identification and authentication process for a service provider. However, current IdMs have not been able to address privacy issues (unauthorised and fine-grained access)that relate to protecting users identity and private data on web services. Many implementations of these frameworks are only concerned with the identification and authentication process of users but not authorisation. They mostly give full control of users digital identities and data to identity and service providers with less or no users participation. This results in a less privacy enhanced solutions that manage users available data in the electronic space. This article proposes a user-centred mandate representation system that empowers resource owners to take full of their digital data; determine and delegate access rights using their mobile phone. Thereby giving users autonomous powers on their resources to grant access to authenticated entities at their will. Our solution is based on the OpenID Connect framework for authorisation service. To evaluate the proposal, we've compared it with some related works and the privacy requirements yardstick outlined in GDPR regulation [1] and [2]. Compared to other systems that use OAuth 2.0 or SAML our solution uses an additional layer of security, where data owner assumes full control over the disclosure of their identity data through an assertion issued from their mobile phones to authorisation server (AS), which in turn issues an access token. This would enable data owners to assert the authenticity of a request, while service providers and requestors also benefit from the correctness and freshness of identity data disclosed to them.

The security of an organization lies not only in physical buildings, but also in its information assets. Safeguarding information assets requires further study to establish optimal security mitigation steps. In determining the appropriate mitigation of information assets, both an information security risk assessment and a clear and measurable rating are required. Most risk management methods do not provide the right focus on ranking the critical information assets of an organization. This paper proposes a framework approach for ranking critical information assets. The proposed framework uses the OCTAVE Allegro method, which focuses on profiling information assets by combining ranking priority measurements using decision support system methods, such as Simple Additive Weighting (SAW) and Analytic Hierarchy Process (AHP). The combined OCTAVE Allegro-SAW and OCTAVE Allegro-AHP methods are expected to better address risk priority as an input to making mitigation decisions for critical information assets. These combinations will help management to avoid missteps in adjusting budget needs allocation or time duration by selecting asset information mitigation using the ranking results of the framework.

The Internet of Things (IoT) is growing rapidly controlling and connecting thousands of devices every day. The increased number of interconnected devices increase the network traffic leading to energy and Quality of Service efficiency problems of the IoT network. Therefore, IoT platforms and networks are susceptible to failures and attacks that have significant economic and security consequences. In this regard, implementing effective secure IoT platforms and networks are valuable for both the industry and society. In this paper, we propose two frameworks that aim to verify a number of security policies related to runtime information of the network and dynamic flow routing paths, respectively. The underlying rationale is to allow the operator of an IoT network in order to have an overall control of the network and to define different policies based on the demands of the network and the use cases (e.g., achieving more secure or faster network).

The wide-spread adoption of system defenses such as the randomization of code, stack, and heap raises the bar for code-reuse attacks. Thus, attackers utilize a scripting engine in target programs like a web browser to prepare the code-reuse chain, e.g., relocate gadget addresses or perform a just-in-time gadget search. However, many types of programs do not provide such an execution context that an attacker can use. Recent advances in data-oriented programming (DOP) explored an orthogonal way to abuse memory corruption vulnerabilities and demonstrated that an attacker can achieve Turing-complete computations without modifying code pointers in applications. As of now, constructing DOP exploits requires a lot of manual work-for every combination of application and payload anew. In this paper, we present novel techniques to automate the process of generating DOP exploits. We implemented a compiler called STEROIDS that leverages these techniques and compiles our high-level language SLANG into low-level DOP data structures driving malicious computations at run time. This enables an attacker to specify her intent in an application-and vulnerability-independent manner to maximize reusability. We demonstrate the effectiveness of our techniques and prototype implementation by specifying four programs of varying complexity in SLANG that calculate the Levenshtein distance, traverse a pointer chain to steal a private key, relocate a ROP chain, and perform a JIT-ROP attack. STEROIDS compiles each of those programs to low-level DOP data structures targeted at five different applications including GStreamer, Wireshark and ProFTPd, which have vastly different vulnerabilities and DOP instances. Ultimately, this shows that our compiler is versatile, can be used for both 32-bit and 64-bit applications, works across bug classes, and enables highly expressive attacks without conventional code-injection or code-reuse techniques in applications lacking a scripting engine.

Efficiently searchable and easily deployable encryption schemes enable an untrusted, legacy service such as a relational database engine to perform searches over encrypted data. The ease with which such schemes can be deployed on top of existing services makes them especially appealing in operational environments where encryption is needed but it is not feasible to replace large infrastructure components like databases or document management systems. Unfortunately all previously known approaches for efficiently searchable and easily deployable encryption are vulnerable to inference attacks where an adversary can use knowledge of the distribution of the data to recover the plaintext with high probability. We present a new efficiently searchable, easily deployable database encryption scheme that is provably secure against inference attacks even when used with real, low-entropy data. We implemented our constructions in Haskell and tested databases up to 10 million records showing our construction properly balances security, deployability and performance.

Industrial Control systems traditionally achieved security by using proprietary protocols to communicate in an isolated environment from the outside. This paradigm is changed with the advent of the Industrial Internet of Things that foresees flexible and interconnected systems. In this contribution, a device acting as a connection between the operational technology network and information technology network is proposed. The device is an intrusion detection system related to legacy systems that is able to collect and reporting data to and from industrial IoT devices. It is based on the common signature based intrusion detection system developed in the information technology domain, however, to cope with the constraints of the operation technology domain, it exploits anomaly based features. Specifically, it is able to analyze the traffic on the network at application layer by mean of deep packet inspection, parsing the information carried by the proprietary protocols. At a later stage, it collect and aggregate data from and to IoT domain. A simple set up is considered to prove the effectiveness of the approach.

This article pertains to cognitive security. YouTube is going to remove videos that deny the Holocaust, other "well-documented violent events, and videos that glorify Nazi ideology or that promote groups that claim superiority to others to justify several forms of discrimination.

The objective of this paper was to propose a 5D combined chaotic system used for image encryption by scrambling and DNA encryption. The initial chaotic values were calculated with a set of equations. The chaotic sequences were used for pixel scrambling, bit scrambling, DNA encryption and DNA complementary function. The average of NPCR, UACI and entropy values of the 6 images used for testing were 99.61, 33.51 and 7.997 respectively. The correlation values obtained for the encrypted image were much lower than the corresponding original image. The histogram of the encrypted image was flat. Based on the theoretical results from the tests performed on the proposed system it can be concluded that the system is suited for practical applications, since it offers high security.

Underwater wireless communication is a critical and challenging research area wherein acoustic signals are used to transfer data. The Underwater Wireless Sensor Network (UWSN) is used to transmit data sensed by the sensors in the sea bed to the surface sinks through intermediate nodes for seismic surveillance, border security and underwater environment monitoring applications. The nodes comprising of UWSN are battery operated and are subjected to failures leading to connectivity loss. And the propagation delay in sending the data in the form of acoustic signals is found to be high and as the depth increases the transmission delay also increases. Hence, routing in UWSN is a complex problem. The simulation experiments of the delay sensitive protocols are found to minimize the delay at the expense of network throughput which is not acceptable. The energy aware routing protocols on the other hand reduces energy consumption and routing overhead but has high delay involved in transmission. In this study, transmission delay and reliability estimation models are developed using which bi-objective routing model is proposed considering both delay and reliability in route selection. In the simulation studies, the bi-objective model reduced delay on an average by 9% and the reliability of the network is improved by 34% when compared to the delay sensitive and reliable routing strategies.

Distributed Denial of Service(DDoS) attacks have become most important network security threat as the number of devices are connected to internet increases exponentially and reaching an attack volume approximately very high compared to other attacks. To make the network safe and flexible a new networking infrastructure such as Software Defined Networking (SDN) has come into effect, which relies on centralized controller and decoupling of control and data plane. However due to it's centralized controller it is prone to DDoS attacks, as it makes the decision of forwarding of packets based on rules installed in switch by OpenFlow protocol. Out of all different DDoS attacks, UDP (User Datagram Protocol) flooding constitute the most in recent years. In this paper, we have proposed an entropy based DDoS detection and rate limiting based mitigation for efficient service delivery. We have evaluated using Mininet as emulator and Ryu as controller by taking switch as OpenVswitch and obtained better result in terms of bandwidth utilization and hit ratio which consume network resources to make denial of service.