Signal amplitude sequence based radio frequency fingerprint feature extraction and identification method

Abstract

The invention discloses a signal amplitude sequence based radio frequency fingerprint feature extraction and identification method. The method includes to detect the locations of initial point and ending point of the received boot transient signal, to collect the boot transient signal sample points, to number the boot transient signal sample points and definite the amplitude functions of the boot transient signal sample points according to the corresponding amplitude of each boot transient signal sample points, to measure the size of the boot transient signal sample points and handle accordingly, to build a sample vector function fingerprint database and an unidentified sample vector function database, to build a between-class dispersion matrix and calculate the trace, and to identify whether the two wireless transmitters are same device. The fingerprint acquisition and feature extraction method is simple and efficient and has general applicability. Meanwhile, by applying the method, the identification rate is obviously improved on the basis of controlling the computing complexity.

Description

technical field [0001] The invention relates to the field of physical layer access authentication of wireless equipment, in particular to a radio frequency fingerprint feature extraction and identification method based on signal amplitude sorting. Background technique [0002] Due to its openness and broadcasting characteristics, wireless communication is extremely vulnerable to security threats such as eavesdropping, forgery and interference. Wireless device access to the network is the first step in wireless communication, access security is the key, and wireless device access security is a serious challenge to wireless network security today. As an important non-password authentication technology, radio frequency fingerprint identification technology has a unique application value in physical layer security because it makes full use of the inimitable radio frequency characteristics to achieve authentication. [0003] Radio frequency fingerprinting is a technology that un...

AI Technical Summary

Problems solved by technology

[0009] The purpose of the present invention is to overcome the deficiencies of the prior art, aim at the extraction and identification of the current radio frequency fingerprint features, provide a radio frequency fingerprint feature extraction and identification method based on signal amplitude sorting, and optimize the existing power-on transient signal sampling method , so that it can take into account the length of the signal along the x-axis, and at the same time, compared with other methods, the computational complexity is significantly reduced, so as to at least improve the correct decision rate, reduce the error probability in the sampling process, and improve the radio frequency fingerprint identification rate the result of

Images