A reconsideration of the pth-order inverse predistorter

The pth-order inverse has been used for nonlinearity compensation for many years. According to pth-order inverse theory, pth-order inverse predistorters and equalizers are identical. However, there is one point often overlooked by many researchers. The pth-order inverse predistorter might change the input signal characteristics so that the Volterra model of the nonlinear system to be compensated, and upon which the pth-order inverse predistorter is designed, can no longer capture the nonlinear system behavior. This phenomenon will degrade the performance of the pth-order inverse predistorter. This performance degradation of a pth-order inverse predistorter becomes more serious when the Volterra model of the nonlinear system is derived for simple testing signals such as PSK. This important, yet often overlooked, phenomenon is investigated in detail