The Metaverse integrates immersive virtual environments, digital interaction, and intelligent computational systems to support new forms of social, educational, industrial, and economic activity. Artificial intelligence is one of the key enabler of this ecosystem because it supports perception, reconstruction, interaction, content generation, personalization, security, and scalable computation. This review presents an AI-centric synthesis of the Metaverse by examining how AI contribute to immersive virtual systems. The paper first outlines the theoretical foundations and enabling technologies of the Metaverse, then reviews major application domains, including intelligent avatars, virtual content creation, recommendation systems, blockchain-enabled digital economies, accessibility, and smart-city digital twins. It further analyzes core AI functions in perception, computation, reconstruction, interaction, and cooperation, followed by a comparative discussion of AI methodologies and their suitability for different Metaverse tasks. The review also discusses public datasets, reproducibility issues, ethical risks, privacy concerns, scalability barriers, and future research directions. Through reducing fragmented discussion and synthesizing methodological trends, this paper provides a structured roadmap for developing trustworthy, scalable, and human-centered AI-enabled Metaverse systems.

With the interconnection of roads spanning not only cities but also multiple countries, transportation capabilities have significantly improved. This has resulted in vehicles being able to reach far-off destinations within shorter timeframes. However, this progress poses challenges in identifying and tracking vehicles, particularly high-security or pursuit vehicles involved in crimes like theft, as they move across various security operational centers. Traditionally, centralized data access through traffic zones has been used for monitoring, surveillance, and tracking purposes. Nevertheless, this approach is susceptible to manipulation and compromises privacy due to unauthorized access by limited stakeholders. To address these concerns, a decentralized alternative is proposed, utilizing a blockchain network that ensures integrity, confidentiality, and access control for vehicle surveillance. Furthermore, computer vision algorithms can be employed for automated surveillance, enabling discreet re-identification or tracking of vehicles. This paper proposes an alternate framework that diverges from the centralized solution, instead encompassing multiple surveillance zones connected through a blockchain-based design with controlled access. To achieve this, we consider traditional approaches to vehicle re-identification using image similarity matching as opposed to resource-intensive unsupervised models performed on a federated consortium blockchain, the Hyperledger Fabric. The framework is evaluated based on latency, resource consumption, and the effectiveness of image processing algorithms for feature sharing through blockchain. Compared to existing frameworks with similar applications, our proposed framework demonstrates higher accuracy and shorter inference times while adhering to the specified constraints. It leverages the concept of blockchain in multi-surveillance zones to enhance privacy and security in automated vehicle surveillance.

Distributed Deep Learning (DDL) partitions deep neural networks across multiple devices, enhancing efficiency in large-scale inference tasks. However, this segmentation exposes intermediate-layer feature maps to new security vulnerabilities, expanding the attack surface beyond traditional input-level threats. This work investigates an adaptation of Natural Evolution Strategies (NES), named NES with Random Uniform Perturbation (NES-RUP), for adversarial manipulation of intermediate-layer feature maps in horizontally distributed inference systems. Instead of Gaussian-based perturbation sampling, the proposed method utilizes uniformly distributed noise and targets only a subset of feature map channels. This design improves stealth by using uniform noise distributions, which avoid extreme outliers and limit perturbations to bounded ranges, keeping activations closer to their clean values and thereby reducing anomaly detection likelihood, while also aligning with the performance and privacy constraints of AIoT-enabled smart environments. Extensive experiments on VGG16, ResNet50 and DeiT-Tiny (a Vision Transformer) using the CIFAR-10 and Mini-ImageNet datasets demonstrate that the adapted NES method achieves high misclassification rates with minimal feature-level distortion, preserving the statistical characteristics of natural feature activations. Furthermore, it successfully bypasses common defenses such as low-pass filtering and feature map anomaly detection (e.g., PseudoNet), revealing critical vulnerabilities in collaborative inference. These findings underscore the need for dedicated defense strategies that address intermediate-layer threats in secure AIoT infrastructures.

The detection of Hardware Trojans (HTs) in electronic components is critical to ensuring the security and integrity of electronic systems, as these malicious modifications can leak sensitive information, alter device functionality, or completely disable the device. This study introduces a novel detection method combining deep learning techniques with dual-channel image transformations to identify HTs via Side-Channel Analysis (SCA). Specifically, our approach converts side-channel time series signals, such as power consumption, electromagnetic emissions, and timing data, into dual-channel, image-like representations using Gramian Angular Field (GAF) and reshaping techniques. These transformations allow convolutional neural networks (CNNs), known for their effectiveness in image analysis, to capture subtle and complex anomalous patterns indicative of HTs. We rigorously evaluated our dual-channel methodology using publicly available Advanced Encryption Standard (AES) datasets for HTs provided by TrustHub and IEEE Dataport. Experimental results demonstrate that the proposed approach achieves superior accuracy compared to existing methods, particularly in challenging datasets. Additionally, we assessed the robustness of our model by introducing varying noise levels to simulate real-world operational perturbations such as process variations, aging, and voltage fluctuations. The proposed method maintains high detection accuracy and demonstrates enhanced resilience under noisy conditions, underscoring its practical applicability and effectiveness in detecting sophisticated HT threats.

Hardware Trojans (HTs) pose a significant threat to the integrity and security of integrated circuits, particularly in critical systems where stealthy hardware modifications can lead to catastrophic consequences. Detecting such Trojans through side-channel analysis (SCA) remains a major challenge due to subtle signal variations and environmental noise. This paper presents a dual-path deep learning framework for the detection and identification of HTs via side-channel analysis. The proposed approach transforms time-series side-channel data- including power consumption, electromagnetic emissions, and timing information-into two distinct image-based representations using Markov Transition Fields (MTF) and a reshaping technique. These transformed representations feed into a two-stage architecture: an Attack Detector, which determines the presence of an HT, and an Attack Identifier, which classifies the specific type of detected HT. To enhance reliability, Monte Carlo Dropout (MCD) is integrated for uncertainty estimation, enabling the framework to flag low-confidence detections and support more robust decision-making. The proposed method is evaluated on publicly available AES hardware Trojan datasets from TrustHub and IEEE Dataport, demonstrating superior accuracy over existing approaches, particularly for complex HT variants with subtle attack signatures. Furthermore, the robustness of the proposed method has been assessed by introducing noise into the validation dataset to simulate real-world operational perturbations such as process variation, aging, and voltage level variations. The results demonstrate the effectiveness and applicability of the dual-path deep learning framework for hardware Trojan detection.

This paper introduces a dual-path deep learning framework specially designed for the efficient detection and classification of Hardware Trojans (HTs) through Side-Channel Analysis (SCA). Using Markov Transition Field (MTF) encoding and a reshaping strategy, the proposed method first converts side-channel time-series signals, including power traces, electromagnetic leaks, and timing data, into two different image-like formats. Detecting subtle Trojan activity requires these representations to capture intricate signal dynamics. Subsequently, each image is processed by a separate convolutional neural network (CNN) branch within the dual-path architecture, with each path optimized for extracting complementary features. To enhance classification performance, the outputs of both CNNs are fused via a dense neural network layer. The dual-path mechanism contributes to improved detection accuracy and robust feature extraction, while the overall architecture supports the classification of specific Trojan types. Evaluation using the publicly available AES Hardware Trojan dataset sourced from TrustHub and IEEE DataPort demonstrates that the proposed model benefits from CNN-based feature learning and outperforms several existing approaches.

The proliferation of Internet of Things (IoT) and Edge devices in critical applications has exposed them to sophisticated Hardware Intrinsic Attacks (HIAs), where hidden trojans can be activated to trigger malicious behavior. Existing intrusion detection systems (IDSs) are largely dependent on network traffic analysis, which struggles to detect hardware-level compromises and evolving threats that mimic normal communication patterns. To overcome these limitations, this paper introduces a novel IDS that leverages hardware-based current profiling for robust intrusion detection. The core of our approach is the monitoring of devices’ current consumption patterns to identify anomalies indicative of an attack, a method that is fundamentally different from and complementary to traditional network-based detection. We demonstrate the efficacy of this approach on a real-world testbed comprising ESP32 microcontrollers (IoT nodes) and a Raspberry Pi 5 (Edge device). The system was evaluated against four critical attack types: Covert Channel Attack (CCA), Power Depletion Attack (PDA), Denial-of-Service (DoS), and Man-in-the-Middle Attack (MIMA). By integrating Artificial Intelligence (AI) through customized Machine Learning (ML) and Deep Learning (DL) models that are trained on over 1.2 million records of current/voltage consumption collected dataset, our IDS achieves a superior multi-class classification with Accuracy of 99.78%, Precision of 99.61%, Recall of 99.61%, and F1-score of 99.60%. These results significantly outperform existing state-of-the-art methods. The findings underscore that physical layer current draw analysis provides a highly reliable and adaptive security mechanism, offering a powerful solution for securing IoT ecosystems against embedded hardware threats and sophisticated cyberattacks.

While Internet of Things (IoT) devices are traditionally perceived as passive network components performing routine tasks with minimal network interaction, this assumption creates a dangerous security blind spot. Malicious actors can exploit manufacturing processes or third-party vendor channels to embed hardware Trojans that remain dormant until triggered to execute programmed attacks. These Hardware Intrinsic Attacks (HIAs) represent an emerging and particularly insidious threat vector that has garnered significant attention in the cybersecurity community. This paper presents an experimental intrusion detection system (IDS) specifically engineered for IoT and Edge computing environments, addressing critical security gaps through hardware-based monitoring approaches. The study evaluates four distinct cyberattack categories: Covert Channel Attacks (CCAs), Power Depletion Attacks (PDAs), Denial-of-Service Attacks (DoS), and Man-in-the-Middle Attacks (MIMAs), systematically implemented across a heterogeneous testbed comprising ESP32 microcontrollers as IoT sensor nodes, Raspberry Pi 5 systems, and Nexys2 Field-Programmable Gate Arrays (FPGAs) as Edge computing platforms. Each attack scenario was meticulously designed to manipulate sensor data integrity, disrupt network communications, or exhaust device resources, accurately simulating real-world IoT security threats. The experimental architecture features multiple ESP32 sensor nodes transmitting data via the UDP protocol to a centralized receiver node connected to a dedicated MATLAB-based monitoring application hosted on a PC for real-time detection analysis. The proposed IDS employs an artificial intelligence (AI) framework utilizing customized machine learning algorithms for sophisticated attack event classification and real-time threat detection. Experimental results demonstrate the superior effectiveness of the proposed IDS architecture in securing IoT and Edge device ecosystems, significantly outperforming existing detection mechanisms.

Unmanned Aerial Vehicles (UAVs) provide many benefits and opportunities across a range of sectors, including surveillance, humanitarian work, disaster management, research, and transportation. Due to their accessibility and affordability, they are now used more than ever, which also poses some challenges. This is the noise pollution produced by the motors and propellers that has been highlighted as a significant issue to the people's health and the environment. To address this issue, this paper proposes to use Generative Adversarial Networks (GAN) to produce an inverse sound signal based on the drone's acoustic signals and use that to cancel the noise produced by the drone. We synthesize training data spanning the acoustic diversity of drone noise: steady-state propeller tones, rapid throttle transitions (simulating ascent/descent), and superimposed broadband turbulence. The GAN model is capable of adapting to dynamic settings, learning from data, and adjusting to testing conditions accordingly. We compared our proposed solution with other techniques that can also be used for drone signal interference in order to suppress the drone noise. This research idea paves the way for the need to address the issue created due to drone noise and a solution in managing this problem for modern drone applications.

Mobile devices have become an essential element in our day-to-day lives. The chances of mobile attacks are rapidly increasing with the growing use of mobile devices. Exploiting vulnerabilities from devices as well as stealing personal information, are the principal targets of the attackers. Researchers are also developing various techniques for detecting and analyzing mobile malware to overcome these issues. As new malware gets introduced frequently by malware developers, it is very challenging to come up with comprehensive algorithms to detect this malware. There are many machine-learning and deep-learning algorithms have been developed by researchers. The accuracy of these models largely depends on the size and quality of the training dataset. Training the model with a diversified dataset is necessary to predict new malware accurately. However, this training process may raise the issue of privacy loss due to the disclosure of sensitive information of the users. Researchers have proposed various techniques to mitigate this issue, such as differential privacy, homomorphic encryption, and federated learning. This survey paper explores the significance of applying federated learning to the mobile operating systems, contrasting traditional machine learning and deep learning approaches for mobile malware detection. We delve into the unique challenges and opportunities of the architecture of in-built mobile operating systems and their implications for user privacy and security. Moreover, we assess the risks associated with federated learning in real-life applications and recommend strategies for developing a secure federated learning framework in the domain of mobile malware detection.

The introduction of smart grids allows utility providers to collect detailed data about consumers, which can be utilized to enhance grid efficiency and reliability. However, this data collection also raises privacy concerns. To protect user privacy, some studies suggest using battery-based load hiding. Nevertheless, the impact of widespread adoption of this approach on utility providers remain unclear. Our paper seeks to evaluate the effects of battery-based load hiding on two critical operations: user profiling and anomaly detection. Our findings reveal that the inclusion of battery users in datasets can diminish the quality of conclusions drawn from these data. This can result in a decrease in the Area Under the Curve (AUC) by more than 10% when attempting to profile users within single-occupant and multiple-occupant households. Furthermore, our experiments demonstrate that battery-based load hiding not only conceals information about users employing the batteries but can also lead to an increased rate of false positives for other non-battery users (from 0.15 to 0.37) within the system. To mitigate these adverse effects, our study assessed various mitigation strategies. In the context of user profiling, our experiment demonstrated that identifying and removing battery users from the analytical dataset using unsupervised detection methods can effectively lessen the impact of battery users. For anomaly detection, our experiment revealed that creating separate classification models for battery and non-battery users can significantly reduce the adverse influence of battery users on the detection performance.

A comprehensive analysis of in vitro pumps used in cardiovascular research is provided in this review, with a focus on the characteristics of generated flows and principles of flow generations. The cardiovascular system, vital for nutrient circulation and waste removal, generates complex hemodynamics critical for endothelial cell function. Cardiovascular diseases (CVDs) could be caused by the disturbances in these flows, including aneurysms, atherosclerosis, and heart defects. In vitro systems simulate hemodynamic conditions on cultured cells in the laboratory to study and evaluate these diseases to advance therapies. Pumps used in these systems can be classified into contact and non-contact types. Contact pumps, such as piston and gear pumps, can generate higher flow rates, but they have a higher risk of contamination due to the direct interaction of pump with the fluid. Non-contact pumps, such as peristaltic and lab-on-disk centrifugal pumps, minimize contamination risks, but they are limited to lower flow rates. Advanced pumps including piezoelectric and I-Cor diagonal pumps are focused on improving the accuracy of flow replication and long-term stability. The operational principles, advantages, and some disadvantages of these pump categories are evaluated in this review, while providing insights for optimizing in vitro cardiovascular models and advancing therapeutic strategies against CVDs. The outcomes of the review elaborate the importance of selecting an appropriate pump system, to accurately replicate cardiovascular flow patterns.

AbstractThe development of smart grids, traditional power grids, and the integration of internet of things devices have resulted in a wealth of data crucial to advancing energy management and efficiency. Nevertheless, public datasets remain limited due to grid operators' and companies' reluctance to disclose proprietary information. The authors present a comprehensive analysis of more than 50 publicly available datasets, organised into three main categories: micro‐ and macro‐consumption data, detailed in‐home consumption data (often referred to as non‐intrusive load monitoring datasets or building data) and grid data. Furthermore, the study underscores future research priorities, such as advancing synthetic data generation, improving data quality and standardisation, and enhancing big data management in smart grids. The aim of the authors is to enable researchers in the smart and power grid a comprehensive reference point to pick suitable and relevant public datasets to evaluate their proposed methods. The provided analysis highlights the importance of following a systematic and standardised approach in evaluating future methods and directs readers to future potential venues of research in the area of smart grid analytics.

Federated Learning (FL) for household-level Short-Term Load Forecasting (STLF) has emerged as a solution to privacy concerns in smart grids, enabling clients to collaboratively train models without sharing their consumption data with a central server. However, sharing model updates can still introduce privacy risks. A common solution to mitigate this risk is the use of differential privacy during the federation process. However, there is a lack of empirical privacy analysis of these techniques in smart grid scenarios. This paper proposes a property inference attack utilizing a single update per client to evaluate privacy risks in federated learning within smart grid contexts. We assess the privacy risk associated with the standard FedAvg algorithm and a differentially private noise-before-aggregation (NBA) FL scheme. Furthermore, we investigate the trade-off between privacy and utility in the NBA-FL scheme. Our empirical findings reveal significant information leakage with standard FedAvg scheme. An adversary with access to a single model update can identify global data properties of the FedAvg client local dataset with an Area Under the Curve (AUC) of 72%. This privacy leakage can be reduced using NBA-FL, which reduces the AUC to 60%. However, the addition of noise to the model updates results in a utility loss of up to 70% in the model’s predictive power. This significant degradation in model performance outweighs the advantages of the scheme.

Drones have become a popular tool for illegal activities and attacks, causing serious threats to global security. In order to address this issue, our project aims to demonstrate the limitations of current drone detection systems by constructing a stealth drone, which is called “Ash.”. The designed drone will be capable of operating in three different modes, which are Wi-Fi, 915 MHz radio frequency (RF) signals, and autonomous mode using a global positioning system (GPS). In addition to that, the drone will be camouflaged to evade detection by optical sensors. We are using long-range (LoRa) technology to transmit on 915 MHz. This makes it difficult to be recognized by the RF analyzer as a drone communication signal. To evade detection by optical sensors, we are camouflaging the drone by adding an air balloon envelope on top of the drone's frame. This makes it appear as a flying air balloon to the detection systems, which should confuse these systems that use computer vision and artificial intelligence. To sum up, this project illustrates the importance of detecting drones accurately and the need for anti-drone systems to adapt to new technologies and tactics. By highlighting the weaknesses of current anti-drone systems, we aim to contribute to the development of more effective technologies to protect global cyberphysical security.

This paper introduces a novel residual-based model to identify households with Battery Electric Vehicles (EVs) under high Air Conditioning (AC) load. The considerable energy demands of AC units can obscure charging events for EVs. In this work we propose a residual based model which leverages the distinctive characteristics of EV charging patterns, marked by unpredictable spikes in energy consumption, and the more predictable nature of AC load. Our proposed approach involves training a lightweight forecasting model to predict overall house-hold consumption and utilizes the residuals of this model for iden-tifying household with EVs. The residual-based model, ResEV-AR, demonstrated a substantial advantage in F1 score (5.8% and 7.32%) compared to state-of-the-art models such as EVS and KBF, respectively. Additionally, a simpler residual model, ResEV-SRM, exhibited a 3.5% F1 score advantage over EVS, coupled with an impressive 11-fold reduction in computation time.

Non-technical loss and energy theft detection are crucial for improving the stability and reducing financial losses in smart grid and power grid utilities. Recently, the availability of massive datasets has improved detection capabilities using sophisticated techniques like deep neural networks. However, training models on extensive feature sets, such as multi-year data, can lead to confusion due to varied behavioral changes in electricity consumption. To address this, we propose a reformulation of the energy theft detection problem by segmenting the time series data and training individual models on each segment. These models’ anomaly scores are then aggregated to produce a final classification. Our framework has shown significant improvement, elevating the F1 score from 0.6 to 0.74, outperforming recent state-of-the-art techniques on the SGCC dataset, the only publicly available dataset labeled for energy theft.

Background: Measuring students' self-regulation skills is essential to understand how they approach their learning tasks in order to identify areas where they might need additional support. Traditionally, self-report questionnaires and think aloud protocols have been used to measure self-regulated learning skills (SRL). However, these methods are based on students' interpretation, so they are prone to potential inaccuracy. Recently, there has been a growing interest in utilizing learning analytics (LA) to capture students' self-regulated learning (SRL) by extracting indicators from their online trace data. Objectives: This paper aims to identify the indicators and metrics employed by previous studies to measure SRL in higher education. Additionally, the study examined how these measurements were validated. Methods: Following the protocol of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), this study conducted an analysis of 25 articles, published between 2015 and 2022, and sourced from major databases. Results and Conclusions: The results showed that previous research used a variety of indicators to capture learners' SRL. Most of these indicators are related to time management skills, such as indicators of engagement, regularity, and anti-procrastination. Furthermore, the study found that the majority of the reviewed studies did not validate the proposed measurements based on any theoretical models. This highlights the importance of fostering closer collaboration between learning analytics and learning science to ensure the extracted indicators accurately represent students' learning processes. Moreover, this collaboration can enhance the validity and reliability of data-driven approaches, ultimately leading to more meaningful and impactful educational interventions.

Study presents a dual-path deep learning architecture designed to effectively detect and identify HT using Side-Channel Analysis (SCA). Initially, the method converts side-channel time series data-including power consumption, electromagnetic emissions, and timing information-into two separate image-like formats by applying Markov Transition Field (MTF) and reshaping techniques. This transformation effectively preserves the complex features of the data, which are crucial for detecting subtle Trojan manipulations. Subsequently, the resulting images are passed through two distinct convolutional neural network (CNN) pathways within our dual-path architecture, each optimized to enhance feature extraction. The features obtained from each pathway are then merged and input into a dense neural network layer, combining various signal characteristics to achieve robust and accurate classification. This dual-path strategy build our proposed method, enabling the simultaneous classification of both legitimate and Trojan-infected data, as well as the identification of the specific type of Trojan attack. The model has undergone extensive testing using the publicly accessible Advanced Encryption Standard (AES) dataset for HTs sourced from TrustHub and IEEE Dataport. Our method not only improves classification accuracy by using the combined strengths of CNNs but also shows marked improvements over current techniques.