Volterra Kernel Extraction Through Monomial Power Series Feed Forward Neural Network for Behavior Modeling of High Speed I/O Buffer

This work explores Volterra kernel extraction using feed forward neural network (FFN) with monomial power series (MPS) activation function for high speed I/O buffer behavioral modeling. The proposed method demonstrates a higher level of accuracy and flexible order truncation. The proposed method is validated using a computer simulated Weiner system with second order nonlinearity. A case study with a real high speed I/O buffer is also conducted and discussed.