Improvement detection system on complex network using hybrid deep belief network and selection features

The challenge for intrusion detection system on internet of things networks (IDS-IoT) as a complex is the constant evolution both large and small attack techniques methods. IoT network growing very rapidly, resulting in data. Complex data produces dimensions one problems IDS networks. In this work, we propose dimensional reduction method to improve performance find out effect IDS-IoT using deep belief (DBN). proposed feature selection uses information gain (IG) principle component analysis (PCA). experiment with DBN successfully detects attacks calculation accuracy, precision, recall, shows that combination PCA superior Wi-Fi datasets. Meanwhile, Xbee dataset PCA. final result measuring average value recall from each IDSDBN test 99%. Other results also show has better than previous studies increasing by 4.12%.