Design and Implementation of Software-Defined Radio Receiver Based on Blind Nonlinear System Identification and Compensation

As the commonly used wideband software radio receiver does not possess the performance of a high linear dynamic range under multiple signal excitation, a new type of adaptive wideband digital receiver architecture is proposed and designed based on blind nonlinear system identification. The traditional narrowband linear receiving and channelizing technologies should not be applied to deal with the complicated multiple-signal excitation with unknown or time-varying characteristics of time domain or frequency domain. Here, the harmonic and the intermodulation components brought about by the wideband digital receiver are firstly identified and extracted in the frequency domain, then a blind identification criterion for minimizing the short-time energy of the nonlinear components is designed, and the steepest descent method (SDM) or the recursive least square (RLS) algorithm is applied to extract and update iteratively the parameters of the nonlinear behavior model of the wideband digital receiver. Finally, the updated model is utilized to compensate the nonlinear distortion of the receiver in real time. The experimental results show that the spur-free dynamic range (SFDR) of the blind identification digital receiver achieves about 20-dB higher than that of the traditional receiver under multitone or 16-QAM bandpass signal excitation. As regards the large bandwidth and the high power efficiency of the RF front end and ADC circuits, the blind identification receiver architecture is helpful for detecting weak signal in concomitance with in-band or out-of-band strong jammers.