Dynamic PA Power Mode DPD on base of 2-Dimensional Chebyshev Polynomials

This paper addresses the challenge of power amplifier (PA) behavioral modeling in scenarios where the PA output power is dynamically varying and non-stationary. Since different PA power modes correspond to distinct nonlinearities, conventional behavioral models-such as the generalized memory polynomial (GMP) and Chebyshev polynomials-require extensions to account for these characteristics. In this work, a two-dimensional Chebyshev polynomial-based model is employed to capture nonlinear distortion across a wide range of PA output power levels, from 0.069 W to 0.912 W, and to predict nonlinear behavior for intermediate power modes. The proposed model achieves high performance, with an adjacent channel leakage ratio (ACLR) of less than −45 dB for the considered PA power modes. Furthermore, the model demonstrates up to a 14 dB improvement in performance compared to a model with the same number of parameters but without PA power mode information.