ANN-LSTM Assisted Intrusion Detection for Next Generation Core Networks

The evolution of smart home automation technologies in Fifth Generation (5G) networks demand for near-instant Internet connectivity to modern conveniences. With so much good coming from technology, it's easy to forget that every item and platform has the potential to be a threat. Despite society's positive opinion of modern technology, cyber security dangers posed by modern technology are a serious hazard. Intrusion Detection System (IDS) is critical in monitoring and detecting intrusion threats in core networks. Traditional firewall approach on data filtering does not detecting all sorts of attacks in real time. This research intends to examine contemporary IDS research using a Deep Learning (DL) methodology, with a focus on datasets, DL methods, and metrics. IDS based on deep learning techniques are particularly good in effectively processing massive volumes of data and identifying any malicious behavior for effective management and prompt detection of these types of attacks. The intrusion detection system's detection rate was improved by focusing on false alarm rate and other performance indicators. The results of studies showed that the DL model LSTM classifier had the best accuracy rate 97%, followed by ANM 95% while the CNN classifier had the lowest accuracy rate 81%.