Abstract
While digital predistortion (DPD) usually targets only the linearity performance of the radio–frequency (RF) power amplifier (PA), this work addresses more than a single PA performance metric exploiting a multi-objective optimization approach. We present a predistorer learning procedure based on a constrained optimization algorithm that maximizes the RF output power, while guaranteeing a prescribed linearity level, i.e., a maximum normalized mean square error (NMSE) or adjacent-channel power ratio (ACPR). Experimental results on a Gallium Nitride (GaN) PA show that the proposed approach outperforms the classical indirect learning architecture (ILA), yet using the same predistorter structure with predetermined nonlinearity and memory orders.
Highlights
Digital predistortion (DPD) is a widely used technique to improve the linearity performance of a power amplifier (PA) at radio–frequency (RF), and it is increasingly becoming a fundamental part of the RF transmitting system
PAs for microwave and millimeter-wave applications are nowadays implemented in Gallium Nitride (GaN) technology, which is affected by spurious dispersive phenomena causing peculiar behaviors like soft compression and long-term memory effects [7,8,9]
The approach has been applied by means of the remote setup in Landin et al [29]. Such a setup is based on a single benchtop instrument, the PXIe-5646R Vector Signal Transceiver (VST) by NI, featuring 200-MHz instantaneous bandwidth
Summary
Digital predistortion (DPD) is a widely used technique to improve the linearity performance of a power amplifier (PA) at radio–frequency (RF), and it is increasingly becoming a fundamental part of the RF transmitting system. The forward model inversion is usually achieved by means of nonlinear optimization algorithms [18] Both DLA and ILA architectures can be implemented using adaptive iterative procedures involving multiple acquisitions, which might allow for the convergence to an improved set of DPD coefficients [20,21]. The first step involves the iterative identification of the predistorted signal in a non-parametric way, meaning that the algorithm does not directly target the DPD coefficients but the optimum input signal realizing the desired output. For all the mentioned architectures, the learning procedure only accounts for one single figure-of-merit (FoM), i.e., a linearity FoM, usually the time-domain normalized mean squared error (NMSE) between the actual and desired output signals These architectures can be generally seen as implementing unconstrained optimization, NMSE minimization being the only target.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
