Optimizing vehicle security: A multiclassification framework using deep transfer learning and metaheuristic-based genetic algorithm optimization

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62690094%3A18450%2F25%3A50022292" target="_blank" >RIV/62690094:18450/25:50022292 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/29142890:_____/25:00052482

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S187454822500006X?via%3Dihub" target="_blank" >https://www.sciencedirect.com/science/article/pii/S187454822500006X?via%3Dihub</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.ijcip.2025.100745" target="_blank" >10.1016/j.ijcip.2025.100745</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Optimizing vehicle security: A multiclassification framework using deep transfer learning and metaheuristic-based genetic algorithm optimization

Popis výsledku v původním jazyce

An extension of the Internet of Things (IoT) paradigm, the Internet of Vehicles (IoV) makes it easier for smart cars to connect to the Internet and communicate with one another. Consumer interest in IoV technology has grown significantly as a result of the increased capabilities of smart vehicles. However, the rapid growth of IoV raises serious privacy and security issues that can lead to dangerous accidents. To detect intrusions into IoT networks, several academics have developed deep learning-based algorithms. Detecting malicious assaults inside vehicle networks and lowering the frequency of smart vehicle accidents are the goals of these models. The proposed approach makes use of an advanced three-layer design that combines ensemble approaches, Genetic Algorithms (GA), and Convolutional Neural Networks (CNNs). Three essential steps are used to execute this methodology: In order to perform CNN-based analysis, we first convert high-level IoV data into image format. The hyperparameters of each base learning model are then optimized via GA, which improves the performance and adaptability of the models. Lastly, we combine the outputs of the three CNN models using ensemble approaches, which greatly improves the intrusion detection system&apos;s (IDS) long-term robustness. Two data sets were used for the evaluations: the CICEVSE dataset, which contains 22,086 samples from 12 distinct intrusion categories, and the publicly accessible Car Hacking dataset, which contains 29,228 samples from five different intrusion categories. According to the experimental findings, the proposed strategy obtained an optimal score of 100% on the Car Hacking images and 93% on the CICEVSE images, demonstrating excellent accuracy. The findings have substantial implications for the development of safe, effective, and flexible intrusion detection systems in the complicated environment of the Internet of Vehicles. © 2025 Elsevier B.V.

Název v anglickém jazyce

Optimizing vehicle security: A multiclassification framework using deep transfer learning and metaheuristic-based genetic algorithm optimization

Popis výsledku anglicky

An extension of the Internet of Things (IoT) paradigm, the Internet of Vehicles (IoV) makes it easier for smart cars to connect to the Internet and communicate with one another. Consumer interest in IoV technology has grown significantly as a result of the increased capabilities of smart vehicles. However, the rapid growth of IoV raises serious privacy and security issues that can lead to dangerous accidents. To detect intrusions into IoT networks, several academics have developed deep learning-based algorithms. Detecting malicious assaults inside vehicle networks and lowering the frequency of smart vehicle accidents are the goals of these models. The proposed approach makes use of an advanced three-layer design that combines ensemble approaches, Genetic Algorithms (GA), and Convolutional Neural Networks (CNNs). Three essential steps are used to execute this methodology: In order to perform CNN-based analysis, we first convert high-level IoV data into image format. The hyperparameters of each base learning model are then optimized via GA, which improves the performance and adaptability of the models. Lastly, we combine the outputs of the three CNN models using ensemble approaches, which greatly improves the intrusion detection system&apos;s (IDS) long-term robustness. Two data sets were used for the evaluations: the CICEVSE dataset, which contains 22,086 samples from 12 distinct intrusion categories, and the publicly accessible Car Hacking dataset, which contains 29,228 samples from five different intrusion categories. According to the experimental findings, the proposed strategy obtained an optimal score of 100% on the Car Hacking images and 93% on the CICEVSE images, demonstrating excellent accuracy. The findings have substantial implications for the development of safe, effective, and flexible intrusion detection systems in the complicated environment of the Internet of Vehicles. © 2025 Elsevier B.V.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

International Journal of Critical Infrastructure Protection

ISSN

1874-5482

e-ISSN

2212-2087

Svazek periodika

49

Číslo periodika v rámci svazku

July

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

17

Strana od-do

"Article number: 100745"

Kód UT WoS článku

001439654300001

EID výsledku v databázi Scopus

2-s2.0-85219149282