Behavioral Power Amplifier Modeling and Digital Predistorter Design with a Chirp Excitation Signal

Digital predistortion techniques are widely utilized to linearize radio frequency power amplifiers (PAs). In their training process, they commonly use filtered Average White Gaussian Noise (AWGN) or the same signal as the transmitted. In this paper an improved calibration waveform for PA behavioral modeling and baseband digital predistortion is proposed. This signal is suitable for different amplifiers, by modifying their configurable parameters. A reduction of the training time and an enhancement in the out-of-band emission level up to 14 dB are achieved.