Duy-Hung Ha

This paper investigates physical layer security (PLS) in a half duplex (HD) relay network with radio frequency energy harvesting (EH). We consider an amplify and forward (AF) protocol in which both the source and the relay harvest energy from a dedicated power beacon, while a passive eavesdropper attempts to intercept the confidential message. Unlike prior studies that rely on asymptotic analysis or numerical integration, we derive exact closed form expressions for the outage probability (OP) and the intercept probability (IP). The proposed expressions characterize the reliability and secrecy trade-off in energy limited cooperative relaying and enable efficient performance evaluation. Monte Carlo simulations validate the analysis and match the theoretical results. Moreover, the closed form results reveal how channel conditions and harvesting duration jointly shape outage and interception behaviors. We further quantify the impacts of normalized transmit power, the time allocation factor between harvesting and information transmission, and the EH efficiency. The results provide practical guidance for selecting system parameters to meet target reliability and secrecy levels, and they show that the time allocation factor must be carefully chosen to balance harvested energy and transmission time.

Phishing detection systems continue to struggle with real-time responsiveness in distributed web environments. This paper proposes a cloud–edge coordination framework that integrates browser-side Uniform Resource Locator (URL) interception, adaptive cache coordination (ACC), and open ONNX-accelerated inference. The lightweight hybrid model (DistilBERT–MLP fusion) fuses DistilBERT embeddings with standardized numerical features through a multilayer perceptron (MLP). Experiments demonstrate a detection accuracy above 99%, millisecond-level neural inference latency, and end-to-end detection times below 150 ms in live browser environments. ACC reduces cold-cache initialization from multisecond delays to microsecond-level cache lookups. Before ACC, a first-time blacklist fetch takes 7–8 s, and subsequent lookups complete in under 1 μs, with total detection latency ranging between approximately 2 and 110 ms. This behavior demonstrates the effectiveness of ACC in reducing repeated network queries. The experimental results substantiate three principal contributions: an efficient hybrid lexical–numeric detector, an ACC-driven mechanism that mitigates cold-start delays, and a cloud–edge design that strengthens real-time phishing protection at the web scale.

This paper explores a radio frequency (RF) energy harvesting nonorthogonal multiple access (NOMA)‐assisted mobile edge computing (MEC) network, consisting of a legitimate unmanned aerial vehicle (UAV), an eavesdropping UAV, and two computing access points (APs) associated with an MEC server. Additionally, one AP is hybrid and provides the UAV with a RF power supply. The NOMA technique enables data offloading to APs over the Nakagami‐ fading channel under imperfect successive interference cancellation conditions. A four‐phase protocol is proposed to meet MEC requirements in terms of latency, security, and energy efficiency. Two key performance indicators are introduced: (i) the naive upper bound of the approximated successful secrecy computation probability and (ii) the naive lower bound of the approximated energy outage probability. Based on analytical and simulation results, a multi‐objective optimization algorithm is applied to determine the system's optimal operating point via the time‐switching ratio, power‐allocation coefficient, and UAV altitude. Monte Carlo simulations validate the theoretical analysis for various system parameters.

In this paper, a novel framework to enhance the reliability of wireless sensor networks (WSNs) by addressing the high probability of outage (OP) resulting from limited energy resources and unreliable channels. The framework integrates three techniques: half-duplex two-way relaying (HD-TWR), digital network coding (DNC), and rateless codes. Although these techniques have been extensively studied in isolation, a comprehensive analysis of their joint performance is provided as the main contribution. The proposed scheme leverages the energy efficiency of HD-TWR, the transmission reduction capability of DNC, and the retransmission-free resilience of rateless codes. Simulation results show that the integrated framework significantly reduces OP, offering a robust and practical solution to enhance the reliability enhancement. Furthermore, the impact of optimal relay node placement is investigated through parameter adjustments in the simulation stage to maximize performance gains.

Conventional analytical models for fault detection in salient-pole synchronous generators often ignore the influence of higher-order force harmonics, limiting their diagnosis accuracy. This paper proposes a harmonic-based analytical model, integrating the Maxwell stress tensor and airgap flux density, and accounting for both radial and tangential components of force under various fault conditions to investigate the impact of inter-turn short circuit (ITSC) and rotor eccentricity (ECC) faults on the force density and vibration characteristics in wound-field synchronous generators (WFSMs). The effects of ITSC and ECC faults are analyzed through spatiotemporal harmonic analysis, showing significant shifts in the harmonic spectrum with additional spatial harmonics at higher frequencies. The proposed analytical model is numerically verified through finite element analysis and experimentally validated on an in-house testbed of healthy and faulty WFSMs. The results prove that the proposed model effectively detects fault-related variations, particularly through higher spatial and temporal harmonic components whereas existing models primarily focus on the fundamental frequency components.

The increasing integration of digital platforms in power systems has enhanced protection capabilities but also introduced new vulnerabilities, particularly in critical components like Line Current Differential Relays (LCDRs). LCDRs offer fast, selective, and sensitive protection for high-voltage transmission lines, but their reliance on synchronized current measurements, communicated over potentially insecure networks, exposes them to cyber threats. Attackers can exploit this dependency through sophisticated attacks, namely False Data Injection Attacks (FDIA) and Time Synchronization Attacks (TSA), manipulating relay decisions and leading to unnecessary tripping, load shedding, and system instability. To address this challenge, this study proposes a two-stage detection and classification framework to distinguish between grid faults and cyber-attacks. The first stage employs an unsupervised Long Short-Term Memory autoencoder, trained solely on normal operational data, to detect anomalous deviations without requiring labeled attack data. Once an anomaly is detected, the second stage, which is based on a supervised Random Forest classifier, categorizes the event as either a physical fault or a specific type of cyberattack. The proposed framework is implemented and validated on the IEEE 14-bus system using the OPAL-RT HYPERSIM platform. Results demonstrate that the proposed framework can accurately detect and classify both faults and cyber-attacks while preserving the essential protective functions of the LCDR.

The perception layer security (PLS) is crucial for ensuring that the data collected by Internet of Things (IoT) devices is accurate, reliable, and protected against various security threats. It helps maintain the overall integrity of the IoT ecosystem and builds trust in its applications. Our work explores the integration of network and PLS in a UAV-enabled mobile edge computing (MEC) system for IoT. This system supports multiple users with a combined non-orthogonal and time-division multiple access (NOTDMA) scheme and is based on backscatter communication (BC). In this system, the UAV-mounted server functions as a hybrid access point (HAP) and hovers over a cluster of energy-constrained IoT devices to transmit RF energy and assist them in performing tasks by employing BC. The IoT devices apply the combined NOTDMA scheme to offload their tasks to the HAP. A mobile passive eavesdropper attempts to intercept information from IoT devices without actively launching any attacks. A partial offloading scheme with various encryption algorithms is proposed to improve the system’s secrecy, which adapts to the users’ non-linear harvested energy levels. In addition, considering the network and physical security, we derive a approximation expression for the secrecy successful computation probability (SSCP). This expression incorporates factors such as harvested energy, local computing and encryption latency, edge offloading latency, processing, decryption, and the associated secrecy costs. The optimization problem for maximizing SSCP is formulated and solved using an Immune algorithm to find the optimal set of device parameters and UAV altitude. Key parameters affecting secrecy and latency performance are analyzed to better understand the system’s behavior. Numerical simulations are provided to validate the accuracy of our analysis.

Detection and isolation of multiple low-severity faults in permanent magnet synchronous machines (PMSMs) under dynamics and noisy conditions are very important to enhance the system’s reliability, lifetime, and service availability. This study investigates a robust fault classification scheme for PMSMs under low-severity, multiple-fault conditions in noisy and variable-speed scenarios. Two supervised learning models, namely Extra Trees (ET) and Support Vector Machine (SVM), are developed and compared using time-domain (TD) and frequency-domain (FD) features. A hybrid feature extraction method is adopted to balance the computational burden and classification accuracy. Hyperparameters and frequency-domain (FD) features are optimized based on prediction confidence, prediction times, and accuracy. Furthermore, the robustness of the data-driven models is evaluated at different Signal-to-Noise (SNR) ratios under mixed faults and variable speeds in an in-house PMSM, combining an inter-turn short circuit fault of 2.2% in one phase and a 30% demagnetization of one rotor magnet. A comparative study was conducted, confirming the proposed ET model’s superior performance in terms of prediction accuracy and robustness.