Neural network based efficiency optimization method for RF Power Amplifiers with controllable power supply

Abstract

Rising energy consumption and thus equipment operating cost as well as the parallel environmental need to reduce the carbon footprint is a very important issue in mobile radio networks. Beside the needs to reduce the energy consumption, the demand for bandwidth and network capacity is increasing and leading to more standards, frequency bands and applications. The RF transceiver, especially the Power Amplifier as a part of it, is one of the bottlenecks and setscrews for the requirements mentioned. The following paper presents a standard and frequency independent concept as well as a method for energy efficiency optimization of RF Power Amplifiers with voltage controllable power supply to adjust the operating point based on Neural Networks. The system is basically standard independent and can adapt itself to a certain standard with its requirements. The system was designed and simulated.