Novel Wireless Fingerprint Extraction and Identification of Power System Terminal Based on Multimodal Features
With the continuous development of a new type of power system,there is a possibility of counterfeit attacks on devices in open environments. A common response is the identification technology based on wireless signal fingerprinting,which ensures the security of the devices at the physical level. However,the instability of wireless signals leads to a low accuracy of identification of terminal devices. Aiming at the above problems,thia paper proposes a novel wireless fingerprint extraction and recognition method for power system terminals based on multimodal features. A dynamic fingerprint database is constructed by fusing the time-domain features and frequency-domain features of wireless signals to form a new type of power system wireless terminal device identification and authentication method based on multimodal features,which realizes stable wireless terminal fingerprint extraction,and can effectively identify and authenticate the wireless devices during wireless device access and wireless device continuous communication to ensure the safe and reliable operation of wireless devices. The experimental results in different scenarios show that the multi-domain fusion method using the frame response time interval as a time-domain feature and the channel state information as a frequency-domain feature for the identification of wireless devices can significantly improve the accuracy and stability of the computation,and the identification rate reaches more than 95%.
Wi-Fi device identificationframe response time intervalchannel state informationdynamic fingerprint database
国家科技期刊平台
NSTL
万方数据
