HNIDD: A Heterogeneous Network Intrusion Detection Approach Using Deep Learning

Nowadays, with the rapid development of computer network technology, heterogeneous networks (HetNets) have been widely adopted in distributed systems. Since HetNets consist of various types of network devices, their core functionality relies on efficient and reliable data transmission. As a result, one of the challenges is the issue of security detection during data transmission, making network intrusion detection (NID) a critical aspect of ensuring data transmission security. This paper presents a HetNet intrusion detection approach using deep learning (DL). We capture network flows data from HetNets, perform feature extraction, and construct a DL model based on convolutional neural networks (CNNs). Additionally, we develop a weighting approach for feature data to optimize experimental results. The results show that our approach achieves an accuracy of $\mathbf{9 9. 1 2 \%}$. Thus, the results indicate that our approach successfully accomplishes intrusion detection in HetNets.