Chebyshev Functional Link Neural Network integrating FIR Filter Architecture for Power Amplifier Linearization

Chebyshev polynomial functional link neural networks (FLNN) integrating FIR filter architecture for power amplifier linearization is proposed. Furthermore, considering the system implementation and resources, we simplify the Chebyshev polynomials in the actual realization on the premise of guaranteeing the fitting accuracy. Experimental results on High-Frequency (HF) Power Amplifier (PA) of actual short-wave communication and the software simulation of dual carrier LTE signal show that more accurate linearization results can be obtained by using the proposed method.