Anomaly-Based Intrusion Detection for IoT Microcontrollers Using Power Side-Channel Fingerprinting
This study introduces a unified evaluation framework for device-level intrusion detection using power side-channel fingerprints under an open-set threat model to target post-enrollment device substitution, and reveals that temporal viability is strongly influenced by feature representation under the evaluated condition...