Detecting Botnet Attacks in IoT Environments Using an Optimized Lightweight Hybrid Deep Learning Model

Abstract

has resulted in billions of interconnected devices operating in smart homes, healthcare systems, industrial automation and intelligent transportation. Although IoT technology provides significant advantages, its limited computational capability and weak security mechanisms expose networks to large-scale cyber threats, particularly botnet attacks. Botnets exploit vulnerable IoT devices and launch coordinated attacks such as distributed denial of service (DDoS), data theft and reconnaissance. Conventional intrusion detection systems based on static rules and signatures fail to effectively detect evolving and previously unseen attacks.

This paper presents an optimized deep learning-based intrusion detection framework for detecting botnet attacks in IoT environments. The framework evaluates existing deep learning models including Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM) and Recurrent Neural Networks (RNN). In addition, a novel lightweight hybrid deep learning model called Advanced Custom Lightweight Recurrent (ACLR) is proposed. The ACLR model integrates Conv1D, LSTM and Simple RNN layers to efficiently capture both spatial and temporal characteristics of network traffic while maintaining low computational complexity.

The Bot-IoT 2018 dataset is used for experimental evaluation. The proposed model achieves an accuracy of 96.42%, outperforming CNN, LSTM and RNN models. The results demonstrate that the proposed framework is effective, scalable and suitable for deployment in resource-constrained IoT environments.