Search Results (1,615)

In real-world applications, user demands for new functionalities and activities constantly evolve, requiring action recognition systems to incrementally incorporate new action classes without retraining from scratch. This class-incremental learning (CIL) paradigm is essential for enabling adaptive and scalable systems that can grow over time. However, Wi-Fi-based indoor action recognition under incremental learning faces two major challenges: catastrophic forgetting of previously learned knowledge and uncontrolled model expansion as new classes are added. To address these issues, we propose CAREC, a class-incremental framework that balances dynamic model expansion with efficient compression. CAREC adopts a multi-branch architecture to incorporate new classes without compromising previously learned features and leverages balanced knowledge distillation to compress the model by 80% while preserving performance. A data replay strategy retains representative samples of old classes, and a super-feature extractor enhances inter-class discrimination. Evaluated on the large-scale XRF55 dataset, CAREC reduces performance degradation by 51.82% over four incremental stages and achieves 67.84% accuracy with only 21.08 M parameters, 20% parameters compared to conventional approaches. Full article

This study presents a lightweight autoencoder-based approach for anomaly detection in digit recognition using federated learning on resource-constrained embedded devices. We implement and evaluate compact autoencoder models on the ESP32-CAM microcontroller, enabling both training and inference directly on the device using 32-bit floating-point arithmetic. The system is trained on a reduced MNIST dataset (1000 resized samples) and evaluated using EMNIST and MNIST-C for anomaly detection. Seven fully connected autoencoder architectures are first evaluated on a PC to explore the impact of model size and batch size on training time and anomaly detection performance. Selected models are then re-implemented in the C programming language and deployed on a single ESP32 device, achieving training times as short as 12 min, inference latency as low as 9 ms, and F1 scores of up to 0.87. Autoencoders are further tested on ten devices in a real-world federated learning experiment using Wi-Fi. We explore non-IID and IID data distribution scenarios: (1) digit-specialized devices and (2) partitioned datasets with varying content and anomaly types. The results show that small unmodified autoencoder models can be effectively trained and evaluated directly on low-power hardware. The best models achieve F1 scores of up to 0.87 in the standard IID setting and 0.86 in the extreme non-IID setting. Despite some clients being trained on corrupted datasets, federated aggregation proves resilient, maintaining high overall performance. The resource analysis shows that more than half of the models and all the training-related allocations fit entirely in internal RAM. These findings confirm the feasibility of local float32 training and collaborative anomaly detection on low-cost hardware, supporting scalable and privacy-preserving edge intelligence. Full article

The rapid growth of Internet of Things (IoT) deployment has led to an unprecedented volume of interconnected, resource-constrained devices. Securing their communication is essential, especially in vehicular environments, where sensitive data exchange requires robust authentication, integrity, and confidentiality guarantees. In this paper, we present an empirical evaluation of TLS (Transport Layer Security)-enhanced MQTT (Message Queuing Telemetry Transport) on low-cost, quad-core Cortex-A72 ARMv8 boards, specifically the Raspberry Pi 4B, commonly used as prototyping platforms for On-Board Units (OBUs) and Road-Side Units (RSUs). Three MQTT entities, namely, the broker, the publisher, and the subscriber, are deployed, utilizing Elliptic Curve Cryptography (ECC) for key exchange and authentication and employing the AES_256_GCM and ChaCha20_Poly1305 ciphers for confidentiality via appropriately selected libraries. We quantify resource consumption in terms of CPU utilization, execution time, energy usage, memory footprint, and goodput across TLS phases, cipher suites, message packaging strategies, and both Ethernet and WiFi interfaces. Our results show that (i) TLS 1.3-enhanced MQTT is feasible on Raspberry Pi 4B devices, though it introduces non-negligible resource overheads; (ii) batching messages into fewer, larger packets reduces transmission cost and latency; and (iii) ChaCha20_Poly1305 outperforms AES_256_GCM, particularly in wireless scenarios, making it the preferred choice for resource- and latency-sensitive V2X applications. These findings provide actionable recommendations for deploying secure MQTT communication on an IoT platform. Full article

This study aims to develop an integrated wireless monitoring system named MANTiS-32, which leverages an open-source platform to enable autonomous modular operation, high-speed large-volume data transmission via Wi-Fi, and the integration of multiple complex sensors. The MANTiS-32 system is composed of ESP32-based MANTiS-32 hubs connected to eight MPU-6050 sensors each via RS485. Four MANTiS-32 hubs transmit data to a main PC through an access point (AP), making the system suitable for real-time monitoring of modal information necessary for structural performance evaluation. The fundamental performance of the developed MANTiS-32 system was validated to demonstrate its effectiveness. The evaluation included assessments of acceleration and frequency response measurement performance, wireless communication capabilities, and real-time data acquisition between the MANTiS-32 hub and the eight connected MPU-6050 sensors. To assess the feasibility of using MANTiS-32 for performance monitoring, a flexible model cable-stayed bridge, representing a mid- to long-span bridge, was designed. The system’s ability to perform real-time monitoring of the dynamic characteristics of the bridge model was confirmed. A total of 26 MPU-6050 sensors were distributed across four MANTiS-32 hubs, and real-time data acquisition was successfully achieved through an AP (ipTIME A3004T) without any bottleneck or synchronization issues between the hubs. Vibration data collected from the model bridge were analyzed in real time to extract dynamic characteristics, such as natural frequencies, mode shapes, and damping ratios. The extracted dynamic characteristics showed a measurement error of less than approximately 1.6%, validating the high-precision performance of the MANTiS-32 wireless monitoring system for real-time structural performance evaluation. Full article

Wi-Fi sensing leverages variations in CSI (channel state information) to infer human activities in a contactless and low-cost manner, with growing applications in smart homes, healthcare, and security. While deep learning has advanced macro-motion sensing tasks, micro-motion sensing such as keypad stroke recognition remains underexplored due to subtle inter-class CSI variations and significant intra-class variance. These challenges make it difficult for existing deep learning models typically relying on fully connected MLPs to accurately recognize keypad inputs. To address the issue, we propose a novel approach that combines a discriminative feature extractor with a Kolmogorov–Arnold Network (KAN)-based classifier. The combined model is trained to reduce intra-class variability by clustering features around class-specific centers. The KAN classifier learns nonlinear spline functions to efficiently delineate the complex decision boundaries between different keypad inputs with fewer parameters. To validate our method, we collect a CSI dataset with low-cost Wi-Fi devices (ESP8266 and Raspberry Pi 4) in a real-world keypad sensing environment. Experimental results verify the effectiveness and practicality of our method for keypad input sensing applications in that it outperforms existing approaches in sensing accuracy while requiring fewer parameters. Full article

The implementation of an Indoor Positioning System (IPS) in a three-storey residential building employing WiFi signals that can also be used to track indoor movements is presented in this study. The movement of inhabitants is monitored through an Android smartphone by detecting the Received Signal Strength Indicator (RSSI) signals from WiFi Anchor Points (APs).Indoor movement is detected through a successive estimation of a target’s multiple positions. Using the K-Nearest Neighbors (KNN) and Particle Swarm Optimization (PSO) algorithms, these RSSI measurements are trained for estimating the position of an indoor target. Additionally, the Density-based Spatial Clustering of Applications with Noise (DBSCAN) has been integrated into the PSO method for removing RSSI-estimated position outliers of the mobile device to further improve indoor position detection and monitoring accuracy. We also employed Time Reversal Resonating Strength (TRRS) as a correlation technique as the third method of localization. Our extensive and rigorous experimentation covers the influence of various weather conditions in indoor detection. Our proposed localization methods have maximum accuracies of 92%, 80%, and 75% for TRRS, KNN, and PSO + DBSCAN, respectively. Each method also has an approximate one-meter deviation, which is a short distance from our targets. Full article

The scope of this paper is the development of a cost-effective wireless torque measurement system for vehicle drivetrain shafts. The prototype integrates strain gauges, an HX711 conditioner, a Wemos D1 Mini ESP8266, and a rechargeable battery directly on the rotating shaft, forming a self-contained sensor node. Calibration against a certified dynamometric wrench confirmed an operating span of ±5–50 N·m. Within this range, the device achieved a mean absolute error of 0.559 N·m. It also maintained precision better than ±2.5 N·m at 95% confidence, while real-time data were transmitted via Wi-Fi. The total component cost is below EUR 30 based on current prices. The novelty of this proof-of-concept implementation demonstrates that reliable, IoT-enabled torque sensing can be realized with low-cost, readily available parts. The paper details assembly, calibration, and deployment procedures, providing a transparent pathway for replication. By aligning with Industry 4.0 requirements for smart, connected equipment, the proposed torque measurement system offers an affordable solution for process monitoring and predictive maintenance in automotive and industrial settings. Full article

Microwave energy is ideal for wearable devices due to its stable wireless power transfer capabilities. However, rigid receiving antennas in conventional RF energy harvesters compromise wearability. This study presents a wearable system using a flexible dual-band antenna (915 MHz/2.45 GHz) fabricated via conformal 3D printing on arm-mimicking curvatures, minimizing bending-induced performance loss. A hybrid microstrip–lumped element rectifier circuit enhances energy conversion efficiency. Tested with commercial 915 MHz transmitters and Wi-Fi routers, the system consistently delivers 3.27–3.31 V within an operational range, enabling continuous power supply for real-time physiological monitoring (e.g., pulse detection) and data transmission. This work demonstrates a practical solution for sustainable energy harvesting in flexible wearables. Full article

This paper presents a dual-band antenna designed for integration into eyewear. The antenna is intended for a system supporting visually impaired individuals, where a wearable camera integrated into glasses transmits data to a remote receiver. To enhance system reliability within indoor environments, the proposed design supports both fifth-generation (5G) wireless communication and Wi-Fi networks. The compact antenna is specifically dimensioned for integration within eyeglass temples and operates in the 3.5 GHz and 5.8 GHz frequency bands. Prototype measurements, conducted using a human head phantom, validate the antenna’s performance. The results demonstrate good impedance matching across the desired frequency bands and a maximum gain of at least 4 dBi in both bands. Full article

Wi-Fi fingerprinting is a widely adopted technique for indoor localization in location-based services (LBS) due to its cost-effectiveness and ease of deployment using existing infrastructure. However, the performance of these systems often suffers due to missing received signal strength indicator (RSSI) measurements, which can arise from complex indoor structures, device limitations, or user mobility, leading to incomplete and unreliable fingerprint data. To address this critical issue, we propose Feature-level Augmentation for Localization (FALoc), a novel framework that enhances Wi-Fi fingerprinting-based localization through targeted feature-level data augmentation. FALoc uniquely models the observation probabilities of RSSI signals by constructing a bipartite graph between reference points and access points, which is then processed by a variational graph auto-encoder (VGAE). Based on these learned probabilities, FALoc intelligently imputes likely missing RSSI values or removes unreliable ones, effectively enriching the training data. We evaluated FALoc using an MLP (Multi-Layer Perceptron)-based localization model on the UJIIndoorLoc and UTSIndoorLoc datasets. The experimental results demonstrate that FALoc significantly improves localization accuracy, achieving mean localization errors of 7.137 m on UJIIndoorLoc and 7.138 m on UTSIndoorLoc, which represent improvements of approximately 12.9% and 8.6% over the respective MLP baselines (8.191 m and 7.808 m), highlighting the efficacy of our approach in handling missing data. Full article

As the number of IoT (Internet of Things) devices continues to grow at an exceptional rate, so does the variety of use cases and operating environments. IoT now plays a crucial role in areas including smart cities, medicine and smart agriculture, where environments vary to include built environments, forest, paddocks and many more. This research examines how Wi-Fi HaLow can be optimised to support the varying environments and a wide variety of applications. Through examining data from performance evaluation testing conducted in varying environments, a framework has been developed. The framework takes inputs relating to the operating environment and application to produce configuration recommendations relating to ideal channel width, MCS (Modulation and Coding Scheme), GI (Guard Interval), antenna selection and distance between communicating devices to provide the optimal performance to support the given use case. The application of the framework is then demonstrated when applied to three various scenarios. This research demonstrates that through the configuration of a number of parameters, Wi-Fi HaLow is a versatile network technology able to support a broad range of IoT use cases. Full article

Human activity recognition (HAR) using Wi-Fi-based sensing has emerged as a powerful, non-intrusive solution for monitoring human behavior in smart environments. Unlike wearable sensor systems that require user compliance, Wi-Fi channel state information (CSI) enables device-free recognition by capturing variations in signal propagation caused by human motion. This makes Wi-Fi sensing highly attractive for ambient healthcare, security, and elderly care applications. However, real-world deployment faces two major challenges: (1) significant cross-subject signal variability due to physical and behavioral differences among individuals, and (2) limited labeled data, which restricts model generalization. To address these sensor-related challenges, we propose TCN-MAML, a novel framework that integrates temporal convolutional networks (TCN) with model-agnostic meta-learning (MAML) for efficient cross-subject adaptation in data-scarce conditions. We evaluate our approach on a public Wi-Fi CSI dataset using a strict cross-subject protocol, where training and testing subjects do not overlap. The proposed TCN-MAML achieves 99.6% accuracy, demonstrating superior generalization and efficiency over baseline methods. Experimental results confirm the framework’s suitability for low-power, real-time HAR systems embedded in IoT sensor networks. Full article

The developed omnidirectional mobile manipulator is an educational omnidirectional mobile manipulator that utilizes the Raspberry Pi Pico W and is programmed in Python. It is designed to enhance STEM education by providing an interactive environment for studying robotics, sensor integration, and programming techniques. The robot is built on an off-the-shelf chassis equipped with Mecanum wheels and a robotic arm actuated by servo motors. As part of this project, the control electronics were designed and implemented to enable seamless operation. While the platform allows students to program the robot as part of the STEM curriculum, our base software solution, developed in Python, provides control of both the mobile base and the robotic arm via a web interface accessible through the robot’s Wi-Fi hotspot. Full article

Robust indoor positioning, crucial to modern location-based services, increasingly leverages Channel State Information (CSI) for its superior multipath resolution over the traditional RSSI. However, current CSI-based methods are hampered by three key limitations: susceptibility to skewed, non-Gaussian noise; informational redundancy from multi-AP configurations; and spatial discontinuities arising from Euclidean-based modeling. To address these challenges, we propose a unified framework that synergistically combines three innovations: (1) an adaptive filtering pipeline that uses wavelet decomposition and dynamic Kalman updates to suppress skewed noise; (2) a graph attention network that optimizes AP selection by modeling spatiotemporal correlations; and (3) a hyperbolic covariance model that captures the intrinsic non-Euclidean geometry of signal propagation. Evaluations on experimental data demonstrate that our framework achieves superior positioning accuracy and environmental robustness over state-of-the-art methods. Crucially, the hyperbolic representation enhances resilience to obstructions by preserving the signal manifold’s true structure, thereby advancing the practical deployment of fingerprinting systems. Full article

The integration of intelligent voice-control systems represents a critical pathway for enhancing driver comfort and reducing cognitive distraction in modern vehicles. Currently, voice assistants capable of accessing real-time vehicular data (e.g., engine parameters) or controlling actuators (e.g., door locks) remain exclusive to premium brands. While aftermarket solutions like Amazon’s Echo Auto provide multimedia functionality, they lack access to critical vehicle systems. To address this gap, we develop a novel architecture leveraging the OBD-II port to enable voice-controlled telematics and actuation in mass-production vehicles. Our system interfaces with a Toyota Hilux (2020) and Mazda CX-3 SUV (2021), utilizing an MCP2515 CAN controller for engine control unit (ECU) communication, an Arduino Nano for data processing, and an ESP01 Wi-Fi module for cloud transmission. The Blynk IoT platform orchestrates data flow and provides user interfaces, while a Voiceflow-programmed Alexa skill enables natural language commands (e.g., “unlock doors”) via Alexa Auto. Experimental validation confirms the successful real-time monitoring of engine variables (coolant temperature, air–fuel ratio, ignition timing) and secure door-lock control. This work demonstrates that high-end vehicle capabilities—previously restricted to luxury segments—can be effectively implemented in series-production automobiles through standardized OBD-II protocols and IoT integration, establishing a scalable framework for next-generation in-vehicle assistants. Full article