Power fingerprinting in SDR & CR integrity assessment

Integrity assessment plays a critical role in the security and regulatory certification of software-defined radio (SDR) and cognitive radio (CR). We propose a novel approach to assess the execution status of SDR and CR by monitoring their dynamic power consumption. The approach relies on extracting distinctive power consumption signatures, or fingerprints, and then using pattern recognition techniques to determine if they match expected behavior. This approach provides an external, independent monitor for assessing the integrity of SDR and CR. Feasibility experiments are encouraging and show the correct identification of basic software routines using discriminatory features extracted from the time domain.