Analysis of even-order terms in memoryless and quasi-memoryless polynomial baseband models

Behavioral modeling of nonlinear passband systems like radio frequency power amplifiers is mainly based on polynomial baseband models. Motivated by the convolution property of the Fourier transform applied to passband signals, it is common practice to include only odd-order terms in these models. Experimental results show, however, that significant improvements can be achieved by also including even-order terms. In this paper, the fundamental relationship of even-order terms in polynomial passband and baseband models is analyzed, providing a theoretical foundation for the improved modeling accuracy of polynomial baseband models with even-order terms.