Cargando...
Fecha
2025-10-01
Derechos de acceso
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editorial
Citas
Resumen
El entorno rural está viviendo una transformación profunda gracias a la incorporación de nuevas tecnologías que permiten mejorar la conectividad y optimizar procesos, especialmente gracias al auge de los dispositivos IoT/IIoT, de ahí la denominación de Smart Rural. Este proceso de digitalización aporta muchas ventajas, pero también introduce nuevas vulnerabilidades y problemas asociados, de ahí que no estén exentos de uno de los principales retos a los que nos enfrentamos hoy en día, la ciberseguridad. El objetivo de este trabajo consiste en abordar esta problemática desarrollando y evaluando RuralShield AI, un IDS adaptativo basado en Deep Learning para analizar tráfico de red en tiempo real y detectar intrusiones en los entornos Smart Rural. La metodología combina el uso de un dataset público como base con la generación de tráfico IoT/IIoT sintético en un simulador de redes, el preprocesamiento de flujos de red, el entrenamiento supervisado de modelos CNN y MLP, la evaluación del rendimiento y su integración en el IDS, desplegado mediante contenedores Docker. Los modelos obtenidos alcanzaron tasas de precisión elevadas y bajos niveles de falsos positivos, demostrando su eficacia frente a los tipos de ataques más representativos del entorno IoT. Además, el proyecto se plantea como una guía metodológica para el reentrenamiento y actualización de modelos, ofreciendo un sistema adaptable frente a las nuevas amenazas de los entornos IoT rurales.
The rural environment is undergoing a profound transformation thanks to the incorporation of new technologies that improve connectivity and optimize processes, especially thanks to the rise of IoT/IIoT devices, hence the name Smart Rural. This digitization process brings many advantages, but it also introduces new vulnerabilities and associated problems, which is why they are not exempt from one of the main challenges we face today: cybersecurity. The objective of this work is to address this problem by developing and evaluating RuralShield AI, an adaptive IDS based on Deep Learning to analyze network traffic in real time and detect intrusions in Smart Rural environments. The methodology combines the use of a public dataset as a basis with the generation of synthetic IoT/IIoT traffic in a network simulator, the preprocessing of network flows, the supervised training of CNN and MLP models, the evaluation of performance, and its integration into the IDS, deployed using Docker containers. The models obtained achieved high accuracy rates and low false positive levels, demonstrating their effectiveness against the most representative types of attacks in the IoT environment. In addition, the project is intended as a methodological guide for retraining and updating models, offering an adaptable system against new threats in rural IoT environments.
The rural environment is undergoing a profound transformation thanks to the incorporation of new technologies that improve connectivity and optimize processes, especially thanks to the rise of IoT/IIoT devices, hence the name Smart Rural. This digitization process brings many advantages, but it also introduces new vulnerabilities and associated problems, which is why they are not exempt from one of the main challenges we face today: cybersecurity. The objective of this work is to address this problem by developing and evaluating RuralShield AI, an adaptive IDS based on Deep Learning to analyze network traffic in real time and detect intrusions in Smart Rural environments. The methodology combines the use of a public dataset as a basis with the generation of synthetic IoT/IIoT traffic in a network simulator, the preprocessing of network flows, the supervised training of CNN and MLP models, the evaluation of performance, and its integration into the IDS, deployed using Docker containers. The models obtained achieved high accuracy rates and low false positive levels, demonstrating their effectiveness against the most representative types of attacks in the IoT environment. In addition, the project is intended as a methodological guide for retraining and updating models, offering an adaptable system against new threats in rural IoT environments.
Descripción
Categorías UNESCO
Palabras clave
IoT, IIoT, IDS, Smart Rural, Ciberseguridad, Tráfico de red, IoT, IIoT, IDS, Smart Rural, Cibersecurity, Network traffic
Citación
Lledó Rovira, Pablo. Trabajo Fin de Máster: RuralShield AI Sistema IDS adaptativo para Entornos Rurales Inteligentes. Universidad Nacional de Educación a Distancia (UNED) 2025
Centro
Escuela Técnica Superior de Ingeniería Informática
Departamento
Sistemas de Comunicación y Control