Very often, resource-constrained IoT devices become the target for DoS or DDoS attacks by intruders that can make the IoT application unavailable to the consumers. For example, consider an IoT-based remote patient-monitoring system. If the sensor's reading of the patient at a critical time, such as during a heart attack, are not available to their doctors or hospital, the patient may lose their life. In this context, devices or host level intrusion detection is essential for most IoT applications. In use case one, we will consider IoT device or host level intrusion detection.
It is essential to select a good feature or set of features to determine anomalies in IoT devices and networks (such as DoS and DDoS) using predictive methods, including DL. Often, we need time series data for real-time or online anomaly detection...