Initial point detection algorithm of transient signal in radio frequency fingerprint identification technology

Abstract

The invention discloses an initial point detection algorithm of a transient part of a radio frequency signal in a radio frequency fingerprint identification technology. The detection algorithm includes the steps of signal acquisition, interception of transient process waveform of each waveform, signal normalization, envelope extraction, envelope filtering, envelope partitioning, Fourier transform of each block of normalized envelope signal, calculation of frequency domain flatness of frequency domain amplitude, time domain flatness and time frequency flatness of each normalized envelop signal, and detection of minimal values of the time and frequency domain flatness as an initial pint of a transient signal. Based on time frequency and frequency domain characteristics, the transient initial point of a radio frequency signal of a transmitter is detected, thereby effectively extracting fingerprint features, and then a classifier is utilized to classify wireless devices, thereby enabling the radio frequency fingerprint identification technology to be better applied to the field of wireless network security. The initial point detection algorithm of the transient part of the radio frequency signal in the radio frequency fingerprint identification technology has a relatively high identification accuracy rate and relatively short classification time.

Description

technical field [0001] The invention relates to a starting point detection algorithm of a radio frequency signal transient part of a radio frequency fingerprint identification technology, and belongs to the field of wireless network security. Background technique [0002] In the process of traditional communication signal detection, the focus of research is often around the acquisition or analysis of communication information transmitted by communication signals, and individual characteristics due to hardware differences are often ignored or filtered out as noise. At present, with the popularization of wireless networks, various security threats to wireless networks are increasing, and the security protocols working on the link layer of wireless networks are often found to have security flaws. For example, there are security flaws in the mutual authentication protocol between RFID readers and tags, which cannot resist the cloning attack of tags. [0003] In order to enhance...

AI Technical Summary

Problems solved by technology

[0004] When the wireless transmitter sends information, the radio frequency signal is divided into a transient part and a steady state part. The detection of the starting point of the transient signal plays a vital role in the correct extraction of the transient signal, because inaccurate detection results will affect the feature extraction. , thereby reducing the classification accuracy

The previously proposed amplitude threshold detection is to compare the amplitude variance of each continuous segment of the signal with the set threshold, which is characterized by low computational complexity, but difficult to choose the threshold

Phase detection is due to the fact that the phase variance is constant in the noise stage, and the phase variance rises when the transmitter is turned on. It is characterized by a small fluctuation of the phase variance and tends to be linear. The threshold is easy to set, but the starting point cannot be clearly detected.

Images