A Power Amplifier Linearization Method for TETRA Based on ANN Digital Predistorter

Baseband digital predistorters (DPDs) are used in wireless communication systems to linearize RF high power amplifiers (HPAs). Traditional digital predistorters based on look-up table (LUT) and polynomial (POLY) either bring in quantization noise or cost large amount of computational resources. In this paper, a baseband digital predistorter based on artificial neural network (ANN) is proposed for terrestrial trunked radio (TETRA). Simulations show that the linearization performance of the proposed DPD is better than the LUT based DPD and the POLY based DPD, and it costs less resources, which means it can be easily implemented.