Deteksi Serangan DDoS pada Intrusion Detection System Berdasarkan Waktu dengan Sliding Window Menggunakan Hybrid 1D CNN-LSTM

With the increasing number of devices connected through the Internet of Things (IoT), personal computers, and mobile devices, the threat of Distributed Denial of Service (DDoS) attacks has become more urgent to address. Such attacks can flood network traffic and disrupt the availability of services in information systems. An Intrusion Detection System (IDS) is one of the key approaches to identifying suspicious activities, including DDoS attacks. This study proposes the implementation of a Hybrid 1D CNN-LSTM method combined with the Sliding Window technique to analyze temporal patterns in the CIC-IDS-2018 dataset. The dataset used is limited to five files with feature selection from 80 attributes, focusing on Benign, DoS, DDoS, and Bot attack types. Through preprocessing steps involving Grey Wolf Optimizer (GWO) feature selection, normalization, and sliding window size 10 segmentation, the proposed model achieved an accuracy of 96%, a macro F1-score of 0.94, a weighted F1-score of 0.96, and an AUC value of 0.99, indicating performance in distinguishing between normal traffic and attacks.