A Notion of Real-Time Anomaly Detection for IoT Devices Based on Hardware-Level Performance

Internet of Things (IoT) is becoming a considerable topic due to its benefits in the modern world. IoT devices carry out simple routine duties, but they can be valuable. IoT devices or a group of devices are connected to the internet, anomaly detection is essential, considering securing the IoT devices within the isolated environments. The most known and typical attacking modes for IoT devices are denial-of-service (DoS) and password brute-force attacks. The most dangerous attack is the Zero-day attack. The best mechanism for finding those issues as a solution is the concept of anomaly detection. Considering IoT device hardware-level anomaly detection mechanism uses the heat and the power consumption for detections. The results of those concepts can be misleading due to environmental situations. Here, it discusses the distinct approach to merely overcoming those problems using CPU and RAM utilization and driving the solution efficiently and effectively up to 99.9%.