Inverse Design of Compact and Wideband Inverted Doherty Power Amplifiers Using Deep Learning

2026-06-25 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, RF & Power Electronics Design · Depth: Expert, quick

Summary

A deep learning-assisted methodology is presented for the inverse synthesis of compact, wideband inverted Doherty power amplifiers (PAs). This approach jointly employs Convolutional Neural Networks (CNNs) and genetic algorithms (GAs) to generate pixelated Doherty combiner networks. These networks integrate load modulation, impedance matching, power combining, and phase compensation into a single structure. As a proof of concept, a GaN HEMT Doherty PA with such a combiner was designed and fabricated. The prototype achieved a measured peak drain efficiency of 51%-63% and a 6-dB back-off efficiency of 48%-54% over 1.9-2.5 GHz. Within this frequency range, the output power was 44+/-0.3 dBm, and with digital predistortion (DPD), the adjacent channel leakage ratio (ACLR) was better than -53.2 dBc.

Key takeaway

For research scientists developing advanced power amplifiers, this methodology offers a path to significantly improve design efficiency. You can use deep learning and genetic algorithms to inversely synthesize compact, wideband Doherty PAs. This integrates complex functions into a single pixelated combiner. The approach enables high efficiencies (51%-63% peak) and output power (44+/-0.3 dBm) across wide frequency ranges (1.9-2.5 GHz). This can accelerate next-generation wireless communication hardware development.

Key insights

Inverse design using deep learning and GAs enables compact, wideband Doherty PA synthesis.

Principles

• Joint CNN and GA use for inverse synthesis.
• Pixelated combiners integrate multiple PA functions.

Method

Convolutional neural networks and genetic algorithms are jointly employed to generate pixelated Doherty combiner networks for inverse synthesis.

In practice

• Design GaN HEMT Doherty PAs.
• Achieve high efficiency over 1.9-2.5 GHz.
• Improve ACLR with DPD application.

Topics

• Inverse Design
• Doherty Power Amplifiers
• Deep Learning
• Genetic Algorithms
• GaN HEMT
• RF Design

Best for: AI Scientist, Research Scientist

Related on AIssential

• Deep Learning-Driven Inverse Design of Doherty Power Amplifiers Using Pixelated Combiners and Dual-State Impedance Synthesis — Deep learning and genetic algorithms enable inverse design of complex Doherty PA combiners for optimal performance.
• Deep-Learning-Based Pixelated Microwave Filter Design and Characterization using Electro-Optical Electric-Field Measurements — Deep learning combined with genetic algorithms automates pixelated microwave filter synthesis, validated by novel electro-optical measurements.
• Inverse Design of Realizable Metasurface based Absorbers using Improved Conditioning and Diversity Enhanced Progressively Growing GANs — A GAN-based inverse design framework synthesizes diverse, physically consistent metasurfaces by integrating EM constraints and diversity regularization.
• Decoupling Numerical and Structural Parameters: An Empirical Study on Adaptive Genetic Algorithms via Deep Reinforcement Learning for the Large-Scale TSP — Dynamic structural adaptation is more critical than numerical fine-tuning for scalable evolutionary algorithms.
• Physics Guided Conditional Diffusion Framework for Generative Inverse Design of Manufacturable Metasurface based Absorbers — Physics-guided diffusion models can rapidly generate manufacturable metasurface designs with high spectral accuracy and multimodal diversity.
• Constructive Circuit Amplification: Improving Math Reasoning in LLMs via Targeted Sub-Network Updates — Targeted sub-network updates can significantly improve specific LLM capabilities with minimal collateral impact.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.