Securing internet of things using machine and deep learning methods: a survey

Cluster Computing, Journal Year: 2024, Volume and Issue: 27(7), P. 9065 - 9089

Published: April 16, 2024

Abstract The Internet of Things (IoT) is a vast network devices with sensors or actuators connected through wired wireless networks. It has transformative effect on integrating technology into people’s daily lives. IoT covers essential areas such as smart cities, homes, and health-based industries. However, security privacy challenges arise the rapid growth applications. Vulnerabilities node spoofing, unauthorized access to data, cyberattacks denial service (DoS), eavesdropping, intrusion detection have emerged significant concerns. Recently, machine learning (ML) deep (DL) methods significantly progressed are robust solutions address these issues in devices. This paper comprehensively reviews research focusing ML/DL approaches. also categorizes recent studies based highlights their opportunities, advantages, limitations. These insights provide potential directions for future challenges.

Language: Английский

---

Overview of AI-Models and Tools in Embedded IIoT Applications

Electronics, Journal Year: 2024, Volume and Issue: 13(12), P. 2322 - 2322

Published: June 13, 2024

The integration of Artificial Intelligence (AI) models in Industrial Internet Things (IIoT) systems has emerged as a pivotal area research, offering unprecedented opportunities for optimizing industrial processes and enhancing operational efficiency. This article presents comprehensive review state-of-the-art AI applied IIoT contexts, with focus on their utilization fault prediction, process optimization, predictive maintenance, product quality control, cybersecurity, machine control. Additionally, we examine the software hardware tools available integrating into embedded platforms, encompassing solutions such Vitis v3.5, TensorFlow Lite Micro v2.14, STM32Cube.AI v9.0, others, along supported high-level frameworks devices. By delving both model applications facilitating deployment low-power devices, this provides holistic understanding AI-enabled practical implications settings.

Language: Английский

---

A Stacking Ensemble Model with Enhanced Feature Selection for Distributed Denial-of-Service Detection in Software-Defined Networks

Engineering Technology & Applied Science Research, Journal Year: 2025, Volume and Issue: 15(1), P. 19232 - 19245

Published: Feb. 2, 2025

The proliferation of Distributed Denial Service (DDoS) attacks poses a significant threat to network accessibility and performance. Traditional feature selection methods struggle with the complexity traffic data, leading poor detection To address this issue, Genetic Algorithm Wrapper Feature Selection (GAWFS) is proposed, integrating Chi-squared (GA) approaches correlation method select most correlated features. GAWFS effectively reduces dimensions, eliminates redundancy, identifies crucial features for classification. Detection accuracy further improved by employing stacking ensemble model, combining Multi-Layer Perceptron (MLP) Support Vector Machine (SVM) as base models, Random Forest (RF) metamodel. proposed classifier achieves impressive accuracies 99.86% training data 98.89% test representing improvements approximately 5% 40%, respectively, over previous studies. time was also reduced 2,593 s, substantial improvement 29.92%. Validation on various benchmark datasets confirmed efficacy approach, underscoring importance enhanced model against DDoS attacks.

Language: Английский

---

Approach for Detecting Attacks on IoT Networks Based on Ensemble Feature Selection and Deep Learning Models

Sensors, Journal Year: 2023, Volume and Issue: 23(17), P. 7342 - 7342

Published: Aug. 23, 2023

The Internet of Things (IoT) has transformed our interaction with technology and introduced security challenges. growing number IoT attacks poses a significant threat to organizations individuals. This paper proposes an approach for detecting on networks using ensemble feature selection deep learning models. Ensemble combines filter techniques such as variance threshold, mutual information, Chi-square, ANOVA, L1-based methods. By leveraging the strengths each technique, is formed by union selected features. However, this operation may overlook redundancy irrelevance, potentially leading larger set. To address this, wrapper algorithm called Recursive Feature Elimination (RFE) applied refine selection. impact set performance Deep Learning (DL) models (CNN, RNN, GRU, LSTM) evaluated IoT-Botnet 2020 dataset, considering detection accuracy, precision, recall, F1-measure, False Positive Rate (FPR). All DL achieved highest F1 measure values, ranging from 97.05% 97.87%, 96.99% 97.95%, 99.80% 99.95%, 98.45% 98.87%, respectively.

Language: Английский

---

A Novel Hybrid Feature Selection with Cascaded LSTM: Enhancing Security in IoT Networks

Wireless Communications and Mobile Computing, Journal Year: 2024, Volume and Issue: 2024, P. 1 - 15

Published: March 13, 2024

The rapid growth of the Internet Things (IoT) has created a situation where huge amount sensitive data is constantly being and sent through many devices, making security top priority. In complex network IoT, detecting intrusions becomes key part strengthening security. Since IoT environments can be easily affected by wide range cyber threats, intrusion detection systems (IDS) are crucial for quickly finding dealing with potential as they happen. IDS datasets have features, from just few to several hundreds or even thousands. Managing such large big challenge, requiring lot computer power leading long processing times. To build an efficient IDS, this article introduces combined feature selection strategy using recursive elimination information gain. Then, cascaded long–short-term memory used improve attack classifications. This method achieved accuracy 98.96% 99.30% on NSL-KDD UNSW-NB15 datasets, respectively, performing binary classification. research provides practical improving effectiveness in networks.

Language: Английский

---

Modeling of Intrusion Detection System Using Double Adaptive Weighting Arithmetic Optimization Algorithm with Deep Learning on Internet of Things Environment

Brazilian Archives of Biology and Technology, Journal Year: 2024, Volume and Issue: 67

Published: Jan. 1, 2024

Language: Английский

---

An efficient deep learning mechanisms for IoT/Non-IoT devices classification and attack detection in SDN-enabled smart environment

Computers & Security, Journal Year: 2024, Volume and Issue: 141, P. 103818 - 103818

Published: March 20, 2024

Language: Английский

---

Optimal Deep-Learning-Based Cyberattack Detection in a Blockchain-Assisted IoT Environment

Mathematics, Journal Year: 2023, Volume and Issue: 11(19), P. 4080 - 4080

Published: Sept. 26, 2023

The Internet of Things (IoT) is the most extensively utilized technology nowadays that simple and has advantage replacing data with other devices by employing cloud or wireless networks. However, cyber-threats cyber-attacks significantly affect smart applications on these IoT platforms. effects intrusions lead to economic physical damage. conventional security approaches are unable handle current problems since threats attacks continuously evolving. In this background, Artificial Intelligence (AI) knowledge, particularly Machine Learning (ML) Deep (DL) solutions, remains key delivering a dynamically improved modern system for next-generation systems. Therefore, manuscript designs Honey Badger Algorithm an Optimal Hybrid Belief Network (HBA-OHDBN) technique cyberattack detection in blockchain (BC)-assisted environment. purpose proposed HBA-OHDBN algorithm lies its accurate recognition classification cyberattacks BC-assisted platform. technique, feature selection using HBA implemented choose optimal set features. For intrusion detection, applies HDBN model. order adjust hyperparameter values model, Dung Beetle Optimization (DBO) utilized. Moreover, BC also applied improve network security. performance was validated benchmark NSLKDD dataset. extensive results indicate model outperforms recent models, maximum accuracy 99.21%.

Language: Английский

---

A Novel ConvXGBoost Method for Detection and Identification of Cyberattacks on Grid-Connected Photovoltaic (PV) Inverter System

Computation, Journal Year: 2025, Volume and Issue: 13(2), P. 33 - 33

Published: Feb. 1, 2025

The integration of solar Photovoltaic (PV) systems into the AC grid poses stability challenges, especially with increasing inverter-based resources. For an efficient operation system, smart grid-forming inverters need to communicate Supervisory Control and Data Acquisition (SCADA) system. However, Internet-of-Things devices that SCADA make these vulnerable. Though many researchers proposed Artificial-Intelligence-based detection strategies, identification location attack is not considered by strategies. To overcome this drawback, paper proposes a novel Convolution extreme gradient boosting (ConvXGBoost) method for only detecting Denial Service (DoS) False Injection (FDI) attacks but also identifying component system was compromised. model compared existing Neural Network (CNN) decision tree (DT) Simulation results demonstrate effectiveness both PV fuel cell (PV-FC) systems. example, accuracy 99.25% 97.76% CNN 99.12% DT during DoS on Moreover, can detect identify faster than other models.

Language: Английский

---

Side-channel attacks on convolutional neural networks based on the hybrid attention mechanism

Deleted Journal, Journal Year: 2025, Volume and Issue: 7(5)

Published: April 24, 2025

Language: Английский

---