Makiyah

The Internet of Things (IoT) has brought convenience to everyday life, particularly through the implementation of smarthome devices. However, the connectivity of these devices also increases their vulnerability to cyber threats, especially Distributed Denial of Service (DDoS) attacks, which can severely disrupt system operations. This study aims to detect DDoS attacks on smarthome devices using the Long Short-Term Memory (LSTM) method, known for its effectiveness in handling sequential data. The dataset used is derived from COMNETS Smarthome, initially in .pcap format and later extracted to .csv using CICFlowMeter. The training process includes several stages: data cleaning, feature selection, label encoding, normalization, and data splitting (training, validation, testing). Evaluation results show that the LSTM model can detect DDoS attacks with a peak accuracy of 99.73%, precision of 99.54%, recall of 100%, and F1-score of 99.77% using an 80:10:10 data split ratio. Therefore, the LSTM model is proven to be effective for DDoS attack detection on smarthome devices and has strong potential to be implemented as an early warning system in IoT networks.