Design of RF PAs Using PSO Algorithm With Dynamic Nonlinear Self-Adaptive Hyperparameters

Abstract

In this work, a new particle swarm optimization (PSO) method with dynamic nonlinear self-adaptive hyperparameters (DNSHs) (PSO-DNSHs) is proposed for radio frequency (RF) power amplifier (PA) design. The theory and advantages of this method are described and analyzed. Based on the proposed PSO-DNSH method, a multioctave PA and a broadband pseudo-Doherty load-modulation balance amplifier (PD-LMBA) are designed. The multioctave single-ended PA achieves an output power exceeding 40.3 dBm, a drain efficiency greater than 60.9%, and a gain of over 10.3 dB between 0.5 and 3.6 GHz. While the broadband PD-LMBA shows a saturated output power of 41.1–43 dBm, a saturation drain efficiency of 57.7%–77.6%, a 6-dB output power back-off (OBO) DE of 47.3%–64.6%, and a 10-dB OBO DE of 51.2%–68.4%. Compared with other existing PSO-based methods, the results demonstrate that the proposed method not only achieves better performance for both PAs, but also shows faster convergence capability. Modulated signal tests along with the digital predistortion (DPD) are also conducted for both PAs, 20-MHz 5G new radio (NR) signal with 6- and 10-dB peak-to-average power ratios (PAPRs) are used for single-ended PA and PD-LMBA, respectively. Excellent linearity is achieved for both PAs.