Behavioral modeling and digital predistortion of power amplifiers with memory using two hidden layers artificial neural networks

This paper presents a novel Two Hidden Layers Neural Networks (2HLANN) model for behavioral modeling and linearization of RF PAs. Starting with a feedback loop principle model of a PA, an appropriate structure is deduced. This structure was then optimized to form a 2HLANN based model capable of predicting the nonlinear behavior and the memory effects of PAs. The validation of the proposed model in mimicking the behavior of a DUT is carried out in terms of its accuracy in predicting the output spectrum, dynamic AM/AM and AM/PM and the NMSE. In addition, the 2HLANN model was used to linearize two 250 Watt PEP Doherty PAs (DPAs) driven with 20 MHz bandwidth signals. The linearization of these DPAs using the 2HLANN enabled attaining an output power of 46.8 dBm and an average efficiency of up to 47.5% coupled with an ACPR higher than 50 dBc. When compared to some published behavioral and DPD schemes, the 2HLANN model demonstrated an excellent modeling accuracy and linearization capability