NFC-IDS: an Intrusion Detection System Based on RF Signals for NFC Security.
Yuwei Yang,Yijie Xun,Yumeng Yan,Jiajia Liu,Ziteng Jin
DOI: https://doi.org/10.1109/iwcmc58020.2023.10182412
2023-01-01
Abstract:The near field communication (NFC), as one of the most widely used radio frequency identification (RFID) technologies, has been applied to intelligent devices to replace the traditional key, bringing convenience to people's lives. While the appearance of NFC keys facilitates users' lifestyles, it also increases the risk of being stolen for intelligent devices. So, it is urgently needed to take action to protect the NFC security of equipment. There are abundant researches on NFC security, which can be divided to protocols authentication and data analysis two main defense methods. However, the way of protocols authentication is limited by the space available and real-time communication of devices. The way of data analysis can not identify the malicious NFC devices from outside. Thus, we propose an intrusion detection system (IDS) based on radio frequency (RF) signals for NFC security, called NFC-IDS. We use the random forests algorithm to select the four most important feature extracted from RF signals and compare random forests, support vector machine (SVM), and k-Nearest Neighbor (k-NN) algorithms to detect intrusions. The experimental results on two real electric motorcycles show that every NFC device has its unique physical signal characteristics, which can be used to detect intrusions with high accuracy and robustness.