A Method for Radio Frequency Model Parameter Estimation based on Improved Intelligent Particle Filter

According to the current application scenario in the field of specific emitter identification, a universal and extensible mathematical model is proposed. This model is utilized to characterize the electronic components that carry radio frequency fingerprints in communication transmitters. An improved intelligent particle filter (IIPF) algorithm is applied to estimate the hardware parameters of the model. Initially, the physical processes of IQ quadrature modulator and power amplifier nonlinear distortion in communication transmitter are modeled in this paper. Subsequently, the transmitted signals are treated as observations, and the particle filter algorithm is used to estimate the state values of the model. To achieve precision in estimation outcomes, an improved intelligent strategy is integrated into the resampling process of particle filtering. The simulation results show that compared with the intelligent particle filter and the traditional particle filter, the mean square error of the improved intelligent particle filter algorithm is significantly reduced in estimating hardware parameters.