“…In our previous works [18], [19], the B-spline neural network has been demonstrated to be very effective in identification and inversion of CV Wiener systems. We adopt two real-valued (RV) B-spline neural networks to model the amplitude response and the phase response of the CV static nonlinearity of the Hammerstein channel, and we develop a highly efficient alternating least squares (ALS) identification algorithm for estimating the channel impulse response (CIR) coefficients as well as the parameters of the two RV B-spline neural networks that model the HPA's CV static nonlinearity.…”