Research On Key Technologies Of Device Identity Authentication Based On Active Fingerprint

As one of the physical layer security enhancement technologies,RF fingerprint authentication technology has the advantages of saving terminal resources,difficult to forge,etc.Therefore,it has attracted extensive attention and research.However,with the advent of the 5G era,the number of mobile communication terminals has exploded,and the problem of indistinguishable RF fingerprint characteristics of similar devices has become more prominent.Aiming at the problem of insufficient discrimination of RF fingerprints,this paper proposes the introduction of active fingerprints at the transmitting end,which takes advantage of the tolerance of the communication protocol itself and magnify the difference in fingerprint characteristics between similar devices.Thus,the network can support more device authentication.This paper studies the corresponding fingerprint extraction and authentication methods to enhance 5G physical layer security based on the uplink sending and receiving process of 5G NR.The main work includes the following aspects:(1)The concept of active fingerprint is introduced and the appropriate RF fingerprint feature is selected as the active fingerprint.The current common RF impairment models such as IQ imbalance,DC offset,CFO,phase noise,and power amplifier nonlinearity,are analyzed.By comparison,IQ imbalance and carrier frequency offset are selected as active fingerprints.(2)For 5G PUSCH,with IQ imbalance and CFO as active fingerprints,a processing scheme of joint estimation and compensation at the receiver is proposed.This method can effectively reduce the influence of active fingerprints on bit error rate while accurately estimating fingerprints.For IQ imbalance,the embedded DMRS is used for channel estimation and equalization to remove the influence of the channel,and then the IQ imbalance are estimated and compensated by the blind compensation algorithm.For CFO,two estimation algorithms,CP-based and pilot-based,are discussed.The simulation results show that these two methods have their own advantages and disadvantages.(3)For 5G PRACH,with IQ imbalance and CFO as active fingerprints,the identification scheme of the receiver is proposed.For CFO,CP-based CFO estimation method and preamblebased CFO estimation method are compared.Simulation results show that the method based on preamble is more accurate.For IQ imbalance,the LMS adaptive filter is used to obtain the channel impulse response,and then the IQ imbalance parameter is estimated by calculation.Simulation results show that the extracted IQ imbalance has the characteristics of high discrimination.Furthermore,it is found that even adding a large active fingerprint has little effect on PRACH detection.(4)A three-dimensional active fingerprint authentication scheme based on PRACH is proposed.The receiver extracts a three-dimensional active fingerprint composed of IQ imbalance parameters and CFO,and establishes an authentication model for legitimate devices through SVDD.The simulation results show that the authentication model has high sensitivity when SNR ≥10d B.The authentication accuracy under various channel scenarios is close to 90%when SNR =5d B,and the accuracy under TDL-D channel is close to 100% when SNR ≥15d B.Finally,this paper proposes an allocation scheme of random selection of active fingerprints.The active fingerprint proposed in this paper can be achieved through the device’s software calibration program or predistortion,and its identification and authentication methods are also applicable to the device’s primitive fingerprint.