Neural Auto-Tuning for Digital Doherty PA Linearization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional analog Doherty Power Amplifiers (DPAs) suffer from inefficiencies and limited operational bandwidth due to their fixed configuration and inability to adapt to varying input signals and environmental changes.
Innovation Solution
A Digital Doherty Power Amplifier (DDPA) system with a learning-based auto-tuning optimization method for the Digital Pre-Distortion (DPD) system, which uses a deep neural network to adaptively find optimal control parameters for various circuit states and input signals, enhancing efficiency and gain while maintaining linearity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional analog DPA is used with fixed configuration, then device complexity is reduced, but adaptability to varying input signals and environmental changes deteriorates
Solution Approach 1:
The patent implements dynamic adaptability by introducing a control circuit that automatically adjusts circuit parameters (such as phase alignment and power splitting ratio) in response to varying input signals and environmental conditions. This transforms the fixed analog DPA into a dynamically adjustable system, resolving the contradiction between adaptability and complexity through automated control mechanisms.
Solution Approach 2:
The control circuit operates autonomously to tune the DPA parameters without requiring manual intervention. The system self-adjusts to optimal operating conditions by monitoring input signals and environmental changes, thereby improving adaptability while keeping the user interface simple and avoiding the need for complex manual tuning procedures.
2Ease of operation
If manual tuning of circuit parameters is performed, then manufacturing precision is improved for fixed operating conditions, but ease of operation deteriorates for varying conditions
Solution Approach 1:
The control circuit incorporates feedback mechanisms that continuously monitor the DPA's operating conditions and automatically adjust parameters to maintain optimal performance. This feedback loop ensures high precision parameter optimization across varying conditions without requiring manual retuning, thereby improving ease of operation while preserving manufacturing precision through automated control.
Solution Approach 2:
The patent replaces manual mechanical tuning procedures with an automated electronic control system. The control circuit electronically adjusts parameters such as phase alignment and power splitting ratio, eliminating the need for physical component adjustments and manual tuning processes. This substitution dramatically improves ease of operation while maintaining or enhancing parameter optimization precision through automated algorithms.
3Adaptability or versatility
If deep learning-based auto-tuning is implemented, then adaptability to various circuit states is improved, but device complexity increases
Solution Approach 1:
The control circuit serves as an intermediary between the input signals and the DPA's internal parameters. It implements deep learning-based algorithms to interpret circuit states and automatically adjust parameters accordingly. This intermediary layer provides sophisticated adaptability while shielding the user from complexity, as the control circuit handles all adjustments autonomously without requiring user understanding of the underlying complex algorithms.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
An auto-tuning controller for improving a performance of a power amplifier system is provided. The controller includes an interface including input terminals and output terminals, the interface being configured to acquire input signal conditions of power amplifiers (PAs), a training circuit including a processor and a memory running and storing a Digital Doherty amplifier (DDA) controller (module), a DPD controller (module) and a DDA-DPD neural network (NN). The training circuit is configured to perform sampling the input signal conditions, and selecting a DPD model from a set of polynomial models for the DPD controller and a set of DDA optimization variables for the DDA controller, using optimized DPD model and DDA coefficients, wherein the optimized DPD model and DDA coefficients are provided by performing an offline optimization for the DPD model and DDA coefficients based on a predetermined optimization method, collecting the optimized DPD coefficients and optimized DDA optimization variables, generating online-DDA optimal coefficients and DPD optimal coefficients using a trained DDA-DPD NN, and updating the generated optimal DDA and DPD coefficients via the output terminals of the interface.