AI-powered biometrics for Internet of Things security: A review and future vision

Journal of Information Security and Applications, Год журнала: 2024, Номер 82, С. 103748 - 103748

Опубликована: Март 21, 2024

Язык: Английский

---

A Machine Learning-Oriented Survey on Tiny Machine Learning

IEEE Access, Год журнала: 2024, Номер 12, С. 23406 - 23426

Опубликована: Янв. 1, 2024

The emergence of Tiny Machine Learning (TinyML) has positively revolutionized the field Artificial Intelligence by promoting joint design resource-constrained IoT hardware devices and their learning-based software architectures. TinyML carries an essential role within fourth fifth industrial revolutions in helping societies, economies, individuals employ effective AI-infused computing technologies (e.g., smart cities, automotive, medical robotics). Given its multidisciplinary nature, been approached from many different angles: this comprehensive survey wishes to provide up-to-date overview focused on all learning algorithms TinyML-based solutions. is based Preferred Reporting Items for Systematic Reviews Meta-Analyses (PRISMA) methodological flow, allowing a systematic complete literature survey. In particular, firstly, we will examine three workflows implementing system, i.e., ML-oriented, HW-oriented, co-design. Secondly, propose taxonomy that covers panorama under lens, examining detail families model optimization design, as well state-of-the-art techniques. Thirdly, present distinct features tools represent current intelligent edge applications. Finally, discuss challenges future directions.

Язык: Английский

---

Noninvasive Diabetes Detection through Human Breath Using TinyML-Powered E-Nose

Sensors, Год журнала: 2024, Номер 24(4), С. 1294 - 1294

Опубликована: Фев. 17, 2024

Volatile organic compounds (VOCs) in exhaled human breath serve as pivotal biomarkers for disease identification and medical diagnostics. In the context of diabetes mellitus, noninvasive detection acetone, a primary biomarker using electronic noses (e-noses), has gained significant attention. However, employing e-noses requires pre-trained algorithms precise detection, often requiring computer with programming environment to classify newly acquired data. This study focuses on development an embedded system integrating Tiny Machine Learning (TinyML) e-nose equipped Metal Oxide Semiconductor (MOS) sensors real-time detection. The encompassed 44 individuals, comprising 22 healthy individuals diagnosed various types mellitus. Test results highlight XGBoost algorithm’s achievement 95% accuracy. Additionally, integration deep learning algorithms, particularly neural networks (DNNs) one-dimensional convolutional network (1D-CNN), yielded efficacy 94.44%. These outcomes underscore potency combining TinyML systems, offering approach mellitus

Язык: Английский

---

Bringing Intelligence to the Edge for Structural Health Monitoring: The Case Study of the Z24 Bridge

IEEE Open Journal of the Industrial Electronics Society, Год журнала: 2024, Номер 5, С. 781 - 794

Опубликована: Янв. 1, 2024

Structural health monitoring (SHM) is key in civil engineering because of the importance and aging infrastructure. We argue that applying leading-edge, data-driven methods large-scale complex industrial systems may be beneficial, particularly for accuracy responsiveness. A fundamental step concerns identification best tools to extract meaningful information from vibrational raw signals. To this end, we study application two convolutional neural network architectures have emerged literature efficient feature extraction time series, namely WaveNet MINImally RandOm Convolutional KErnel Transform (MiniRocket). The test bench Z24 bridge progressive damage classification dataset. Results show a model based on reaches state-of-the-art performance, also reducing size computational complexity. proves perfectly suited interpret vibration waveforms directly domain, without any specific preprocessing. On other hand, MiniRocket excels ease configuration (only hyperparameters are tweaked), overall training efficiency, size, lending itself as valuable agile alternative (e.g., rapid prototyping). Our main advancement is, thus, characterization highly effective methods, employable different SHM tasks. assessed performance models embedded platforms, proposing smart sensor system where local hub collects signals constellation inertial sensors infers assessment onsite, allowing self-assess its state resorting connectivity nor cloud resources.

Язык: Английский

---

Design and Implementation of an Internet-of-Things-Enabled Smart Meter and Smart Plug for Home-Energy-Management System

Electronics, Год журнала: 2023, Номер 12(19), С. 4041 - 4041

Опубликована: Сен. 26, 2023

The demand response program is an important feature of the smart grid. It attempts to reduce peak demand, improve grid efficiency, and ensure system reliability. Implementing demand-response programs in residential commercial buildings requires use meters plugs. In this paper, we propose architecture for a home-energy-management based on fog-computing paradigm, Internet-of-Things-enabled plug, meter. plug measures real-time root mean square (RMS) value current, frequency, power factor, active power, reactive power. These readings are subsequently transmitted meter through Zigbee network. Tiny machine learning algorithms used at identify appliances automatically. were prototyped by using Raspberry Pi Arduino, respectively. plug’s accuracy was quantified comparing it laboratory measurements. To assess speed precision small algorithm, publicly accessible dataset utilized. obtained results indicate that both exceeds 97% 99%, execution trained decision tree support vector verified 3 Model B Rev 1.2, operating clock 600 MHz. measured latency classifier’s inference 1.59 microseconds. practical situation, time-of-use-based can cost about 30%.

Язык: Английский

---

Tiny Machine Learning (TinyML) for Efficient Channel Selection in LoRaWAN

IEEE Internet of Things Journal, Год журнала: 2024, Номер 11(19), С. 30714 - 30724

Опубликована: Июнь 12, 2024

Язык: Английский

---

Machine Learning With Computer Networks: Techniques, Datasets, and Models

IEEE Access, Год журнала: 2024, Номер 12, С. 54673 - 54720

Опубликована: Янв. 1, 2024

Machine learning has found many applications in network contexts.These include solving optimisation problems and managing operations.Conversely, networks are essential for facilitating machine training inference, whether performed centrally or a distributed fashion.To conduct rigorous research this area, researchers must have comprehensive understanding of fundamental techniques, specific frameworks, access to relevant datasets.Additionally, data can serve as benchmark springboard further investigation.All these techniques summarized article; serving primer paper hopefully providing an efficient start anybody doing regarding using learning.

Язык: Английский

---

TinyIDS - An IoT Intrusion Detection System by Tiny Machine Learning

Lecture notes in computer science, Год журнала: 2024, Номер unknown, С. 71 - 82

Опубликована: Янв. 1, 2024

Язык: Английский

---

Device-Level Energy Efficient Strategies in Machine Type Communications: Power, Processing, Sensing, and RF Perspectives

IEEE Open Journal of the Communications Society, Год журнала: 2024, Номер 5, С. 5054 - 5087

Опубликована: Янв. 1, 2024

Язык: Английский

---

Advancements in TinyML: Applications, Limitations, and Impact on IoT Devices

Electronics, Год журнала: 2024, Номер 13(17), С. 3562 - 3562

Опубликована: Сен. 8, 2024

Artificial Intelligence (AI) and Machine Learning (ML) have experienced rapid growth in both industry academia. However, the current ML AI models demand significant computing processing power to achieve desired accuracy results, often restricting their use high-capability devices. With advancements embedded system technology substantial development Internet of Things (IoT) industry, there is a growing desire integrate techniques into resource-constrained systems for ubiquitous intelligence. This aspiration has led emergence TinyML, specialized approach that enables deployment on resource-constrained, power-efficient, low-cost Despite its potential, implementation such devices presents challenges, including optimization, capacity, reliability, maintenance. article delves TinyML model, exploring background, tools support it, applications advanced technologies. By understanding these aspects, we can better appreciate how transforming landscape IoT systems.

Язык: Английский

---