Agnostic Envelope Linearization of Dynamically Supplied Power Amplifiers for Mobile Terminals

Area of Science:

• Electrical Engineering
• Wireless Communications
• Signal Processing

Background:

• Envelope tracking (ET) power amplifiers (PAs) are crucial for 5G new radio (NR) mobile terminals.
• Nonlinear distortion in ET PAs degrades signal quality and impacts adjacent channel power ratio (ACPR).
• Existing linearization techniques often require output monitoring, increasing system complexity.

Purpose of the Study:

• To propose a novel envelope linearization technique, envelope optimization (EOPT), for ET PAs in 5G NR mobile terminals.
• To address nonlinear distortion without requiring output monitoring at the envelope tracking modulator (ETM).
• To validate the EOPT method's performance and efficiency on a system-on-chip (SoC) ET PA.

Main Methods:

• The EOPT method utilizes an envelope generalized memory polynomial (EGMP) behavioral model for linearization.
• An iterative nonlinear least squares solution, inspired by the deep deterministic policy gradient (DDPG) algorithm, is employed to extract EGMP model coefficients.
• The technique is validated on a dedicated SoC ET PA board using 5G NR test signals.

Main Results:

• The EOPT method successfully compensated for nonlinear distortion in the ET PA.
• Achieved adjacent channel power ratio (ACPR) below -36 dBc with 16.8% power efficiency for 60 MHz 5G NR signals at 2.55 GHz.
• Linearization performance is comparable to envelope leakage cancellation (ELC) but avoids the need for an analog-to-digital converter (ADC).

Conclusions:

• The proposed EOPT method effectively linearizes ET PAs for 5G NR mobile terminals.
• The technique offers a power-efficient solution that meets stringent linearity requirements.
• EOPT provides a competitive alternative to existing linearization methods, simplifying system design by eliminating the need for ETM output monitoring.