This study investigates Distributed Denial-of-Service (DDoS) attack detection within smart home environments using both traditional machine learning and deep learning approaches. Real smart home traffic data, collected approximately 11.5 h of normal and attack activity, was used to implement and evaluate two models: k-Nearest Neighbour (k-NN) and an Artificial Neural Network (ANN). The k-NN model achieved an accuracy of 97.13%, while the ANN achieved 81.7% accuracy under the same dataset conditions. Unlike previous studies relying solely on benchmark datasets, this work uses self-collected smart home data to assess model feasibility and real-world deployment potential.
Raja et al. (Sat,) studied this question.