Intrusion Detection System For IoT Networks Using Neural Networks With Extended Kalman Filter

Although coined in 1999, Internet of Things (IoT) has been one of the most sought-after technologies since the 1980s. However, with its remarkable growth comes the need to protect it against highly sophisticated cyber-attacks. Acknowledging the fact that such attacks are not totally avoidable, early detection becomes essential. Over the years, the field of Machine Learning has been explored in detecting various network-based attacks. Hence, we take it into our account for creating an intelligent Intrusion Detection System (IDS) for IoT networks using a Neural Network with Extended Kalman Filter (EKF). The proposed system has been evaluated using two datasets, NSL-KDD and BoT-IoT datasets. The proposed system has been analyzed using several metrics such as accuracy, detection rate, and false negative rates.