CLAE-MLP: A Deep Learning Framework for Botnet Detection in IoT Network Using N-BaIoT Dataset

Abstract

Modern industries and day-to-day activities have experienced major progress from the Internet of Things because it links devices instantly to share real-time data. Internet of Things devices connecting to each other creates exposure to cyber-attacks and produces botnet attacks that damage network security while spreading data breaches and blocking service access. To protect IoT networks you must detect and stop botnet threats. This research creates a deep learning model that combines CNN autoencoders with LSTMs and MLPs to better find botnet attacks in IoT networks. The CNN module generates spatial understanding of data patterns while the LSTM module finds temporal sequences. The MLP module helps make better predictions and reduces incorrect findings. Our system can check network traffic quickly to find botnets that protect digital security. Experimental tests prove that this model finds both familiar and new security threats to strengthen IoT protection.