“‘latex This paper introduces ”SmartEdge,” a novel approach for optimizing the processing and analysis of Internet of Medical Things (IoMT) data. Through systematic preprocessing, advanced feature selection, controlled sampling, and innovative model aggregation, SmartEdge enhances performance and efficiency while minimizing computational overhead. The continuous improvement loop ensures adaptability to new data and operational conditions, promising significant advancements in IoMT ecosystem management. Our feature engineering on the CICIoMT2024 dataset emphasizes dimensionality reduction to improve computational efficiency in edge IoMT security and cyber attack detection. Techniques like Principal Component Analysis (PCA), feature selection, and embedding methods reduce the dataset’s dimensionality by 95%, drastically decreasing computational load and enabling real-time processing on resource-limited edge devices. This streamlining facilitates faster and more efficient machine learning model deployment, significantly boosting cyber attack detection and prevention in IoMT environments. Focusing on cyber resilience, we trained 10 edge models in just 55 seconds, with near-zero aggregation time due to an efficient method requiring no additional training or predictions. Using a small data set, five top features, and PCA with two components, we achieved high-performance results, with each model and the global model reaching accuracy above 0.99.

The European Digital Identity Wallet (EUDI Wallet) is the flagship initiative under the novel eIDAS1 regulation, aiming at enhancing digital identification by providing all EU citizens with a universal and secure digital ID. This ID will facilitate a wide range of activities, including travel, work, cross-border education, access to online services, payments, document signing, and more. Current efforts concentrate on specifications for Person Identification Data (PID) and other attribute attestations via Verifiable Credentials (VC) and the requirements for PID Providers, Qualified and Electronic Attribute Authorities (QAA and EAA), EUDI Wallet Providers, Relying Parties, and other actors within the EUDI Wallet ecosystem. This paper delves into further essential requirements of the eIDAS 2.01 regulation and proposes the FIDO framework as a solution for central aspects such as strong customer authentication, user sole control and device attestation. FIDO is not only an efficient framework for Strong User Authentication (SUA), but has also been recognized by the EU Cybersecurity Agency (ENISA)2 and the ETSI TR 119 460 report3 as compliant with eIDAS eID schemes. In this paper, we first briefly state the requirements of the EUDI Wallet Architecture and Reference Framework (ARF) for online and proximity identification and authentication, and propose leveraging the FIDO framework for authentication in the EUDI Wallet (section 1). Next, we highlight the potential complementarity of the current ARF design effort and the FIDO framework, explaining how the use of FIDO can achieve economies of standardization and accelerate the adoption of the EUDI Wallet by service providers that rely on additional security guarantees for LoA High, such as financial institutions and critical infrastructure operators (section 2). In section 3 we outline the technical details for seamlessly integrating a FIDO authenticator into the EUDI Wallet architecture. Finally, we introduce the need for protected confirmation, as a means to enable EUDI Wallet users to “reliably control what they confirm on their device”, and recommend to also leverage protected confirmation for selective disclosure (section 4).

The increasing need to protect individual privacy in data releases has led to significant advancements in privacy-preserving technologies. Differential Privacy (DP) offers robust privacy guarantees but often at the expense of data utility. On the other hand, data pooling, while improving utility, lacks formal privacy assurances. Our study introduces a novel hybrid method, termed PoolDiv, which combines differential privacy with data pooling to enhance both privacy guarantees and data utility. Through extensive simulations and real data analysis, we assess the performance of synthetic datasets generated via traditional DP methods, data pooling, and our proposed PoolDiv method, demonstrating the advantages of our hybrid approach in maintaining data utility while ensuring privacy.

Unmanned Aerial Vehicles (UAVs) have become indispensable assets in various sectors, leveraging their mobility and data collection capabilities. However, privacy and security concerns have fueled interest in Federated Learning (FL) as a solution. FL, decentralized and collaborative, offers promise in addressing privacy risks inherent in centralized data processing while enhancing model performance. In this review, we explore FL’s privacy and security implications in UAV ecosystems. We highlight FL’s potential to mitigate privacy risks by aggregating model updates locally, minimizing data transmission needs. Additionally, we examine security challenges and evaluate protective mechanisms. Through a systematic literature review, we identify gaps and propose future research directions, aiming to enhance the security and privacy of FL in UAV applications.

Aiming to address issues such as low security, inefficient encryption and decryption, weak robustness, image tampering, and the misuse of existing image encryption algorithms in blockchain-based applications for image sharing and transmission, this paper proposes a traceable image privacy protection scheme based on blockchain and homomorphic algorithm. Firstly, Paillier algorithm and image block technology were used to preprocess and encrypt the image. Feature images are extracted from encrypted images using lifting wavelet transform (LWT) and singular value decomposition (SVD). By means of dynamic S-box, binary watermark information is subjected to confusion, diffusion and XOR operations to generate zero-watermarking images for lossless embedding of traceability information. Then, the zero-watermarking image and its corresponding key are securely stored on the International File System (IPFS) through a smart contract mechanism to regulate image transactions and enable traceability of the image. Finally, the smart contract is utilized to detect any inaccuracies and errors in the zero-watermarking image, enabling tamper detection of said image. In experimental tests, the proposed scheme was shown to enable data distribution and computation with private data in ciphertext state, exhibiting high efficiency in encryption and decryption. Additionally, the zero-watermarking algorithm exhibits strong anti-attack capabilities and robustness. By considering image privacy protection and secure storage, tamper detection and traceability of encrypted images are effectively addressed.

In digital forensics, proper evidence management is crucial for police investigations due to its volatility and improper handling technique. However, to present and utilize the captured digital forensic images, which are typically concrete and admissible in court and a trial, is very challenging. In addition, centralized evidence-gathering, classification, and archiving techniques diminish the credibility of digital evidence. To address these issues, this work proposed a blockchain-based classified digital forensic sensitive image management framework. The framework includes a rapidly expanding polygon chain, Interplanetary File System (IPFS), and a deep learning model that performs with more than 98% accuracy. This framework preserves sensitive forensic images in the fully decentralized blockchain. Our system uses metamask wallet-based authentication to guard against unauthorized users. We use IPFS to produce the image hash to encrypt and save the forensic image in the blockchain. A voting method based on smart contracts is suggested for accessing the preserved image. We implemented the framework, which offers promising performances in response time, evidence insertion time, evidence access time, communication overhead, and effective forensic image categorization.

Modern technologies of computing cloud are showing great promise, but at the same time create new security challenges that hinder full acceptance. Given that most of these services often use cloud networks as channels for communication, securing data transmission is crucial. This paper proposes a hybrid encryption algorithm, the proposed two-layered PRC6 cipher, developed to achieve high security in cloud computing environments with minimal resource constraints. The PRC6 cipher incorporates enhancements from Cha-cha into an extension of the RC6 cipher. PRC6 implements double encryption. At the first level, the plain text is divided into four equal parts, each encrypted by processes derived from RC6, which include shifting, summation, modulo arithmetic, and XOR with a generated key. The second level incorporates a Quarter round function, among others, to further obscure the encoded message. PRC6 is implemented in a parallel computing model to significantly reduce overall computation time, especially important for lightweight applications. Experimental results show that the algorithm can achieve a high level of security for cloud workloads in only a few encryption rounds. Performance evaluations against popular encryption standards also indicate that PRC6 offers promising security benefits when computational resources are limited. This hybrid approach presents a viable solution for strengthening data protection in modern cloud systems.

A document by K. Ramesh. Click on the document to view its contents.

Phishing attacks are thoroughly engineered attacks where the attackers use emails, messages, and websites of reputed sources as a medium to trick their targets into sharing sensitive content. This sensitive content primarily consists of their financial information, in the case of small attacks whereas some planned advanced attacks also target to obtain their login information. In the past few years, there has been a noticeable shift in attackers’ priorities, moving away from targeting individuals and instead concentrating on the organization’s employees. It is also observed that most cyber-attacks are the result of employee negligence. Due to the widespread availability of phishing kits and the expansion of ransomware as a service (RaaS), aspiring hackers now possess a straightforward method to defraud individuals. What is particularly worrisome about this growing trend is that individuals lacking technical expertise are engaging in such activities using simple tools and online instructional materials. Machine learning can help in recognizing different phishing attacks and patterns. We describe several classical algorithms to detect Phishing attacks. We aim to utilize machine learning techniques like Multilayer perceptron, Random Forest, XG Boost, and different classifiers for detection. This paper will compare various studies for detecting phishing attacks using each Artificial Intelligence technique: Deep Learning and Machine learning. In order to enhance our study, we have also listed various other conventional methods of detection that do not utilize the benefits of training through machine learning models.

In the ever-evolving landscape of cybersecurity, the application of game theoretic models has emerged as a powerful and innovative approach to enhance our understanding and management of cyber threats. This abstract explores the application of a variant of game theoretic models within the context of a Cyber Threat Intelligence (CTI) framework. With the proliferation of cyber-attacks targeting critical infrastructure, sensitive data, and national security, it has become imperative to develop proactive and adaptive strategies for threat detection, mitigation, and response. The variant of game theoretic models discussed in this abstract departs from traditional game theory by incorporating elements of dynamic adaptation and machine learning. This adaptation enables the framework to model and analyze the intricate and rapidly changing interactions between threat actors and defenders in real-time, thereby providing a more accurate representation of the evolving threat landscape. By leveraging machine learning algorithms, the model can continuously learn and adapt to new threats and tactics, making it a versatile tool for CTI. This abstract also explores the practical applications of the variant model in various aspects of cybersecurity, including threat actor profiling, vulnerability assessment, and decision support for incident response. By considering the strategic motivations and behaviors of threat actors, organizations can make informed decisions regarding resource allocation, risk assessment, and security investments. The integration of this variant of game theoretic models into CTI holds great potential to revolutionize our approach to cybersecurity, enabling organizations to stay one step ahead of adversaries. As the digital world becomes increasingly complex, the ability to predict, mitigate, and adapt to cyber threats is crucial for safeguarding critical assets and ensuring the resilience of digital infrastructure. This paper highlights the significance of this innovative approach and its potential to shape the future of cyber threat intelligence and cybersecurity practices.

In the realm of social media, cyberbullying’s pervasive impact raises urgent concerns about its emotional and psychological toll on victims. This study addresses the imperative of effectively detecting cyberbullying. By leveraging ML and DL techniques, we aim to develop reliable methods that accurately identify instances of cyberbullying in social media data, enhancing detection efficiency and accuracy. This facilitates timely intervention and support for affected individuals. In this comprehensive analysis of existing systems, various ML and DL models are extensively texted for cyberbullying detection. The evaluated models include Random Forest, XgBoost, Naive Bayes, SVM, CNN, RNN, and BERT. Pre-processed datasets are utilized to train and evaluate the models. To evaluate the ability of each model to reliably identify cyberbullying in social media data, performance metrics such as F1 score, recall, precision, and accuracy are used. The findings of this study demonstrate the efficacy of different ML and DL models in monitoring cyberbullying in social media data. Among the models evaluated, the BERT model exhibits exceptional performance, achieving the highest accuracy rates of 88 .8% for binary classification and 86 .6% for multiclass classification.

This article looks at Healthcare and the issues that exist with current cybersecurity measures. As attacks, including those related to ransomware become more commonplace it is important to provide safeguards to protect the data of patients and healthcare organizations. Examples of cybersecurity breaches are looked at with insights shared on how they happened and what could have been done to prevent them. Rather than the original perimeter security approach which many organizations took a newer approach known as Zero-Trust looks at how to neutralize attacks even if a hacker has penetrated the perimeter security. Key areas involved in Zero-Trust are reviewed and explained as to how the applicability to healthcare can make a difference in protecting individuals and healthcare organizations.

Ransomware is a menace to the vibrant digital ecosystem. The exponential growth in ransomware attacks, its detrimental impacts, and the ever-changing methods adopted by threat actor groups demands a focused understanding of the evolution of ransomware. This would help the organizations devise novel defensive frameworks and security controls against the modern ransomware. In this work, the impacts and evolution of ransomware through different phases up to its current form are detailed. Further, based on the study and analysis of the most prevalent modern ransomware variants, their most used tactics, techniques and procedures (TTPs) are identified as per the MITRE ATT&CK model. This acts as a platform to propose a generic attack model for ‘modern ransomware’. Building on the existing MITRE mitigation, D3FEND-based approaches and considering the resource and budget constraints of organizations, a simplified three-tier defensive model that is cost-effective and implementable is put forward. Thus, this work aims to open avenues for understanding the TTPs, and attack methodology of ‘modern ransomware’, thereby developing feasible and implementable defensive security controls.