Internet of Things (IoT) devices are increasingly susceptible to botnet threats, with Mirai-based attacks posing significant security challenges. To address these concerns, we present a lightweight anomaly detection model based on a Gaussian Mixture Model (GMM), tailored for detecting Mirai botnet traffic. By leveraging anomaly detection techniques, our approach identifies malicious activity with optimized latency and computational efficiency, making it suitable for resource-constrained IoT environments. Trained specifically on Mirai-related traffic, the proposed model achieves comparable detection accuracy while significantly improving latency compared to traditional decision tree-based models. These results underscore the effectiveness of the GMM approach in balancing detection performance and real-time responsiveness for IoT applications.

The growing complexity of autonomous railway systems poses significant challenges for predictive maintenance, often resulting in unexpected failures, increased costs, and safety risks. This study proposes an advanced method utilizing the XGBoost algorithm to detect failures early in train air production units (APUs). By integrating sophisticated feature engineering with robust machine learning, the proposed method identifies anomalies up to three days in advance, enabling proactive maintenance interventions. An integrated SHAP analysis clarifies the model’s decision-making process, enhancing interpretability and fostering trust in predictions. Experimental evaluations demonstrate exceptional performance, achieving precision, recall, and F1-Score of 99.95%. The model’s robustness is validated by a Mean Cross-Validation Accuracy of 99.95% (±0.003%) and a Test Set Accuracy of 99.95%, surpassing prior studies. These findings highlight the proposed method’s potential to significantly enhance the safety, resilience, and efficiency of autonomous railway systems.

Recent advances in Predictive Maintenance (PdM) are revolutionizing the railway industry by enabling early detection of failures and optimizing maintenance strategies. This research introduces a novel approach for predicting the Remaining Useful Life (RUL) of train Air Production Units (APUs) using advanced deep learning models. The proposed methodology leverages a Bidirectional Long Short-Term Memory (BiLSTM) network to analyze complex temporal patterns in sensor data, significantly enhancing RUL prediction accuracy. A key innovation of this work lies in the integration of the MetroPT and MetroPT-3 datasets with maintenance reports, creating a comprehensive dataset for prediction. Prior to model development, data processing and feature selection were performed using a GradientBoostingRegressor. Extensive testing demonstrates the model’s exceptional performance, achieving a Mean Absolute Error (MAE) of 0.08, a Mean Squared Error (MSE) of 0.01, and a Root Mean Square Error (RMSE) of 0.11. These results underscore the approach’s effectiveness in accurately predicting the RUL of train APUs. Additionally, SHAP analysis was conducted to enhance the model’s interpretability by elucidating its predictions and highlighting key feature interactions, further validating its reliability.

The Internet of Things (IoT) demands reliable communication between resource-constrained devices. Selecting appropriate routing protocols is crucial for achieving this goal, but traditional methods struggle with the unique challenges of IoT networks, including security, mobility, energy limitations, and scarce resources. This paper proposes a novel routing protocol that addresses these challenges by focusing on two key parameters: energy and Signal-to-Interference-plus-Noise Ratio (SINR). Our protocol excludes devices with low remaining energy from the chosen path, while prioritizing links with strong SINR. This strategy ensures the discovery of stable and energy-efficient paths for extended network lifetime. Furthermore, we introduce an enhanced security layer. During route discovery, intermediate devices exchange secret keys using Weil Pairing and Smart-Chen-Kudla schemes, guaranteeing both confidentiality and mutual authentication while minimizing computational overhead. Extensive simulations on Network Simulator 2 (NS-2) demonstrate significant improvements in throughput and end-to-end delay compared to existing solutions.

This paper proposes a dynamic policy based access control model for the Internet of Things (IoT) enabled smart healthcare. Traditional access control models including Mandatory Access Control (MAC), Discretionary Access Control (DAC), Role-based Access Control (RBAC), and Attribute-based Access Control (ABAC) are based on static policies. This makes policy management very heavy to implement in modern healthcare systems where context changing is highly dynamic, requiring constant policy updating. To overcome the gaps of existing access control models, the proposed solution introduces a dynamic ABAC approach which makes the access policy dynamically adaptive to satisfy emergency and real-time monitoring requirements. Evaluation results demonstrate the effectiveness of the proposed approach.

Electric vehicles (EVs) have gained significant attention due to their environmental and energy-efficient benefits. Dynamic Wireless Charging (DWC) has emerged as one of the most promising EV charging methods because it overcomes the challenges associated with traditional charging stations. However, as EVs keep moving while charging, this requires specific authentication and payment methods, enabling EV mobility privacy and ensuring fair billing. To address these concerns, we propose in this paper a smart contract based solution to deal with privacy and fair billing for EVs dynamic wireless charging. The proposed solution includes a lightweight authentication mechanism enabling a continuous authenticate-and-charge process when EV is moving on the road.

In-motion wireless charging services have emerged for the development of Electric Vehicles (EVs) industry. Fast authentication and privacy awareness are the major concerns in this promising technology. This paper proposes FLPA, a fast and anonymous authentication scheme for EVs charging on the move, which preserves the identity and location privacy of EVs during the recharging process. In order to deal with EV's high mobility and constrained resources of charging pads, FLPA provides lightweight and fast EVs authentication to charging pads while ensuring secure and fair-billing based on authenticated pairwise keys and coin chains. Through a performance analysis, we demonstrate our scheme's advantage compared to current solutions.

Abstract Securing Mobile ad hoc Networks (MANET) is a challenging task, notably due to the lack of an online infrastructure. In particular, key management in MANET is a problem for which many solutions have been proposed in literature. Unfortunately, these solutions are rather limited in terms of security and availability of keys. In this paper, we propose a secure, robust, and fully distributed scheme for public‐key certificate management in MANET. Our scheme, based on threshold cryptography, ensures that the private key of the certificate authority will not be revealed to an adversary, even if the number of compromised shareholders exceeds the threshold of vulnerability, thereby thwarting mobile‐adversary attacks. We describe SRKM in detail and, by using security analysis and simulations, show its effectiveness, robustness and security. Copyright © 2010 John Wiley & Sons, Ltd.

This paper proposes robust key management scheme (RKM ): a new certification management scheme for mobile ad-hoc networks. RKM, based on threshold cryptography, ensures better robustness compared to other schemes proposed in the literature. In particular, it guarantees the confidentiality of the private key of the certification authority, even if the number of compromised nodes exceeds the threshold of vulnerability. Thus, our model offers better protection against many attacks such as mobile adversary attack.