Deep Learning-Based Intrusion Detection for IoT Networks: A Scalable and Efficient Approach DOI Creative Commons
Md. Alamgir Hossain

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Abstract The rapid expansion of the Internet Things (IoT) has revolutionized industries by enabling seamless connectivity, but it also introduced significant security vulnerabilities, making IoT networks prime targets for cyberattacks. Traditional intrusion detection systems often struggle to cope with high volume and dynamic nature traffic, necessitating development more robust intelligent mechanisms. This research presents a deep learning-based approach real-time threat in networks, leveraging advanced models such as 1D Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) Recurrent (RNNs), Multi-Layer Perceptrons (MLPs) enhance detection. study utilizes CIC IoT-DIAD 2024 dataset, comprehensive collection flow-based network traffic containing both benign attack scenarios. proposed were trained evaluated on feature sets, optimizing hyperparameters maximize accuracy, recall, F1-score. In multi-class classification, CNN achieved highest accuracy 99.12%, followed LSTM (98.98%), RNN (98.43%), MLP (97.21%). For binary anomaly detection, again demonstrated superior performance an 99.53%, while LSTM, RNN, 99.52%, 99.25%, 98.78%, respectively. results indicate that is most effective model excelling extraction classification. findings contribute scalable efficient solutions, improving ability detect mitigate cyber threats environments.

Language: Английский

Deep Learning-Based Intrusion Detection for IoT Networks: A Scalable and Efficient Approach DOI Creative Commons
Md. Alamgir Hossain

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Abstract The rapid expansion of the Internet Things (IoT) has revolutionized industries by enabling seamless connectivity, but it also introduced significant security vulnerabilities, making IoT networks prime targets for cyberattacks. Traditional intrusion detection systems often struggle to cope with high volume and dynamic nature traffic, necessitating development more robust intelligent mechanisms. This research presents a deep learning-based approach real-time threat in networks, leveraging advanced models such as 1D Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) Recurrent (RNNs), Multi-Layer Perceptrons (MLPs) enhance detection. study utilizes CIC IoT-DIAD 2024 dataset, comprehensive collection flow-based network traffic containing both benign attack scenarios. proposed were trained evaluated on feature sets, optimizing hyperparameters maximize accuracy, recall, F1-score. In multi-class classification, CNN achieved highest accuracy 99.12%, followed LSTM (98.98%), RNN (98.43%), MLP (97.21%). For binary anomaly detection, again demonstrated superior performance an 99.53%, while LSTM, RNN, 99.52%, 99.25%, 98.78%, respectively. results indicate that is most effective model excelling extraction classification. findings contribute scalable efficient solutions, improving ability detect mitigate cyber threats environments.

Language: Английский

Citations

0