276 research outputs found
Recent advances in the application of data science to industrial cybersecurity [Avances recientes en la aplicación de la ciencia de datos a la ciberseguridad industrial]
No full textGuest Editorial: Advanced Computing and Blockchain Applications for Critical Industrial IoT
No full textDeveloping Secure Cloud Applications
No full textToday the main limit to Cloud adoption is related to the perception of a security loss the users have. Indeed, the existing solutions to provide security are mainly focused on Cloud service provider prospective in order to securely integrate frameworks and Infrastructures as a Services in a Cloud datacenter. Customer could not monitor and evaluate the security mechanisms enforced by service provider.
Service Level Agreements mainly focus on performance related terms and no guarantees are given for security mechanisms. Customers are interested in tools to verify and monitor the implemented security requirements. On the other hand, developers need tools to deploy Cloud applications offering measurable security grants to end users. In this paper, we propose an approach to implement security mechanisms as components in the application design process. We modeled security interactions according to the specific threat, the specific security requirements and user/application capabilities trying to improve security. It enables a Service Provider to offer security guarantees to customers.
The approach has been designed to fit with different Cloud platforms, but to demonstrate its applicability, we will present a case study on the mOSAIC Platform
Targeting Gene Function in Germinal Center B Cells: A Practical Approach
No full textThe germinal center (GC) reaction represents an essential phase of an adaptive immune response. Dysfunction of GC B cells can lead to life-threatening diseases including autoimmune disorders, lymphomas, and opportunistic infections. Defining the molecular circuitries controlling GC B cell physiology is crucial to understand the pathogenesis of GC B cell disorders, as well as to develop improved vaccines against foreign pathogens. Conditional gene targeting based on the Cre/loxP recombination system has substantially accelerated our comprehension of the genetic networks controlling GC B cell function. Several independent studies in the past 10 years have highlighted the many advantages and the few limitations and pitfalls associated to conditional gene manipulation in GC B cells using the Cre/loxP recombination system. Here, we describe the basic features of GC B cell-specific gene targeting experiments. We provide indications on the type of Cre transgene and controls to be chosen, way-out strategies to overcome leakiness of the Cre/loxP system, and approaches to minimize the number of experimental animals and to speed up analyses on conditional mutant GC B cells
Ineffectiveness of Digital Transformations for Detecting Adversarial Attacks Against Quantized and Approximate CNNs
No full textConvolutional Neural Networks (CNNs) have achieved superhuman performance in computer vision tasks. However, these networks are becoming both increasingly complex and resource-intensive, and are susceptible to adversarial attacks. On one hand, to counter complexity and resource-related limitations, various techniques such as Quantization and Approximate Computing (AxC) have been proposed to reduce the complexity and power consumption of CNNs, respectively. On the other hand, various techniques have been proposed to craft more precise and stronger adversarial attacks, as well as new methodologies to defend against them. Nevertheless, the relationship between the efficiency and security of CNNs is not adequately addressed. Therefore, this article examines the potential for detecting adversarial attacks against CNNs through image transformation, in the context of quantized and approximate neural networks. The experimental results indicate that image-transformation techniques are not effective in detecting adversarial samples against quantized and approximated CNNs, despite their success in detecting such samples against floating-point CNNs
Ensuring End-to-End Security in Computing Continuum Exploiting Physical Unclonable Functions
No full textIn recent years, there has been an increase in Cloud Continuum adoption to support Internet of Things applications. Inevitably, such a paradigm introduces novel security challenges, particularly concerning the security of communicating nodes to prevent malicious actors from tampering within the network, and ensuring the confidentiality of sensitive data during transmissions. Traditional security methods often fall short in addressing these issues, especially where network nodes are built upon resource-constrained devices. Consequently, the scientific community has begun exploring the potential of Physical Unclonable Functions (PUFs), which are unique digital identifiers derived from the inherent variability in the manufacturing process of integrated circuits, as a means to enhance security mechanisms at minimal overhead cost. This paper introduces Secure-PHEMAP (S-PHEMAP), a novel and lightweight PUF-based key management scheme designed for end-to-end communications that guarantees authenticity, confidentiality and integrity for pair communications. The proposed scheme builds upon the PHEMAP protocols, inheriting its security properties. S-PHEMAP can be employed in scenarios where both communicating devices embeds a PUF or in situations where only one of them has a PUF. In addition, the paper includes a deployment strategy in a Cloud Continuum domain, by leveraging the Chef automation framework
- …
