Behavioral Modeling and Predistortion

The innovative technique, digital predistortion is the linearization technique widely used for applications with or up to 20 MHz bandwidth. The performance of behavioral modeling is influenced by two key aspects, the observation and the formulation. The observation refers to accurate acquisition of the signals at the input and output of the device under test (DUT) and formulation refers to choice of a suitable mathematical relation that describes all the significant interaction between the DUT's input and output signals. The approaches that can be considered to synthesize the predistortion function consists of synthesizing and then inverting the model of the DUT to be linearized. The second approach consists of identifying the digital predistortion function directly. The behavior of the DUT can be experimentally measured using the instantaneous complex baseband waveform approach. The formulations proposed for behavioral modeling and digital predistortion include the nested look-up table, the Volterra model, memory polynomial, envelope memory polynomial, and, Wiener model.