Intrusion detection systems (IDS) stand as formidable guardians in network security, playing a pivotal role in identifying and mitigating potential threats. As our digital landscape evolves, the imperative for robust intrusion detection mechanisms grows exponentially. The significance of IDS extends beyond preserving the integrity, confidentiality, and availability of network resources. In the dynamic realm of evolving cyber threats, IDS acts as the frontline defender-constantly monitoring network traffic to pinpoint suspicious activities and preemptively mitigate security breaches. In this paper, we study the effectiveness of combining the potency of the transformer-based model known as Bidirectional Encoder Representations from Transformers (BERT) in conjunction with the Synthetic Minority Over-sampling Technique (SMOTE) and a Multi-Layer Perceptron (MLP) to enhance the classification tasks in Machine Learning (ML) based IDS. We tested our approach on well-known and recent datasets, demonstrating that it is possible to obtain very high accuracy and robust performance in the detection of various attack types even when the datasets are affected by class imbalance. Beyond these results, our research introduces a novel perspective by seamlessly integrating the interpretability and context awareness of BERT with the efficient classification of MLP. This novel approach holds promise for advancing intrusion detection capabilities and contributing to the broader cybersecurity community.

The rapidly evolving landscape of cyber threats poses significant challenges to network security, particularly in decentralized environments such as edge computing. This paper proposes an enhanced Intrusion Detection System (IDS) architecture that integrates Transfer Learning (TL) to create a unified supermodel, enabling adaptability and scalability for detecting emerging threats across diverse datasets. The key contributions of this study include: (1) leveraging BERT-based feature extraction to enhance the semantic understanding of intrusion patterns, (2) employing an MLP classifier refined through TL to improve classification performance (3) addressing class imbalance using Synthetic Minority Over-Sampling Technique (SMOTE), and (4) optimizing model deployment by distributing the heaviest computational tasks for the creation of the unified supermodel across the edge nodes with the available capacity, thereby reducing latency and enabling accurate real-time threat detection by resource constrained IoT devices. The proposed TL-enabled supermodel is periodically updated and shared with the IoT devices, ensuring robust and adaptive security mechanisms without the need for extensive local training. Our experimental evaluation on CIC-IDS 2017 and NSL-KDD 2009 datasets demonstrates the effectiveness of the approach, achieving 99% accuracy, precision, recall, and F1-score. Our results highlight the scalability, efficiency, and real-world applicability of our IDS framework, reinforcing its role in fortifying network security within highly dynamic cyber threat landscapes.

While technological advancements and their deep integration in connected and automated vehicles is a central aspect in the evolving trend of automotive industry, they also depict a growing size attack surface for malicious actors: the latter ones typically aim at exploiting known and unknown security vulnerabilities, with potentially disastrous consequences on the safety of vehicles, people, and infrastructures. In recent years, remarkable efforts have been spent to mitigate security vulnerabilities in intelligent and connected vehicles, in particular in the inside of vehicles, the so-called intra-vehicle networks. Despite those efforts, securing intra-vehicle networks remains a non-trivial task due to their heterogeneous and increasingly complex context. Starting from the above remarks and motivated by the industrial research and innovation project EMERGE, in this paper we report on a novel cryptographic hardware-software solution that we have designed and developed for securing the intra-vehicle network of intelligent connected vehicles: the Crypto-Engine. The Crypto-Engine relies on a lightweight hybrid-key cryptographic scheme to provide confidentiality and authentication without compromising the normal communication performance. We tested the Crypto-Engine and demonstrated that, once configured according to application-defined performance requirements, it can authenticate parties and secure the communications with a negligible overhead.

Advances in technology call for a parallel evolution in the software. New techniques are needed to support this dynamism, to track and guide its evolution process. This applies especially in the field of embedded systems, and certainly in Wireless Sensor Networks (WSNs), where hardware platforms and software environments change very quickly. Commonly, operating systems play a key role in the development process of any application. The most used operating system in WSNs is TinyOS, currently at its TinyOS 2.1.2 version. The evolution from TinyOS 1.x and TinyOS 2.x made the applications developed on TinyOS 1.x obsolete. In other words, these applications are not compatible out-of-the-box with TinyOS 2.x and require a porting action. In this paper, we discuss on the porting of embedded system (i.e., Wireless Sensor Networks) applications in response to operating systems’ evolution. In particular, using a model-based approach, we report the porting we did of Agilla, a Mobile-Agent Middleware (MAMW) for WSNs, on TinyOS 2.x, which we refer to as Agilla 2. We also provide a comparative analysis about the characteristics of Agilla 2 versus Agilla. The proposed Agilla 2 is compatible with TinyOS 2.x, has full capabilities and provides new features, as shown by the maintainability and performance measurement presented in this paper. An additional valuable result is the architectural modeling of Agilla and Agilla 2, missing before, which extends its documentation and improves its maintainability.