Feature extraction from I/Q signals for ZigBee devices identification

Abstract

In this letter, we have developed a novel and simple feature extraction that exploits the characteristics of I/Q signals received from ZigBee devices to represent RF circuit impairments such as the frequency offset and I/Q imbalance. Experiments have been carried out to evaluate the identification performance in several scenarios: line-of-sight (LOS) and non line-of-sight (NLOS) environments, varying distance between the transmitter and receiver. The results show that our method can identify twelve ZigBee devices with high accuracy. In addition, it is revealed that some features are sensitive to environmental change during the training and test phases.