Journal of Applied Sciences ›› 2019, Vol. 37 ›› Issue (1): 12-23.doi: 10.3969/j.issn.0255-8297.2019.01.002

Previous Articles     Next Articles

Estimation of Broadband Communication Transmitter Fingerprints Based on B-Spline Neural Network

YUAN Hong-lin, LU Xiao-dan, XU Chen   

  1. School of Electronics and Information, Nantong University, Nantong 226019, Jiangsu Province, China
  • Received:2018-05-07 Revised:2018-10-17 Online:2019-01-31 Published:2019-01-31

Abstract:

A method for the identity authentication of orthogonal frequency division multiplexing (OFDM) communication devices with the received signal is proposed. The IQ imbalance and nonlinearity of the transmitter are estimated as the transmitter fingerprints. Firstly, the multipath channel impulse response is estimated according to the conjugate symmetric pilot. Secondly, the channel impulse response estimation, the conjugate antisymmetric pilot, and the linear approximation of the nonlinear power amplifier are used to estimate the IQ imbalance parameter combination of the transmitter. Then, the B-Spline neural network model coefficients of the nonlinearity of the transmitter are estimated. Finally, the similarity factor of the nonlinear model coefficient estimation is extracted, which constructs the feature vector of the transmitter fingerprint with the estimated IQ imbalance parameter combination. The feature vector is used to recognize or confirm the identity of the communication devices. Theoretical derivation and numerical experiments show that the proposed method can be applied to the physical layer high-intensity authentication and anti-counterfeiting of OFDM communication devices.

Key words: physical layer authentication, IQ imbalance, B-Spline neural network, communication transmitter fingerprints, radio frequency (RF) fingerprints, nonlinearity

CLC Number: