Digital Predistortion Model Synthesis from RF Power Amplifier Circuits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital predistortion (DPD) technologies struggle to simultaneously achieve high accuracy and low complexity in modeling RF power amplifier systems, especially in 5G systems with GaN devices, due to the lack of a physical model that accounts for thermal and trapping effects, and the mismatch between circuit-level and DPD models.
Innovation Solution
A method and system for forward synthesis of a digital predistortion nonlinear model derived from a circuit description, involving the transformation of equivalent circuit models into baseband-discrete time-domain equations through integro-differential equation systems, Laplace transforms, and z-transforms to create an accurate and efficient DPD model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a black box model is used for DPD, then the model is simple to implement, but the modeling accuracy deteriorates when dealing with sophisticated power amplifier circuit structures
Solution Approach 1:
The patent introduces an intermediate physical model that bridges the gap between black box models and full physical models. This intermediate model incorporates essential physical characteristics (thermal effects, trapping effects) while maintaining mathematical tractability for DPD applications, serving as a mediator that balances simplicity and accuracy.
Solution Approach 2:
The patent transforms the full physical model parameters into DPD-appropriate parameters through systematic transformation. By changing the representation form while preserving physical meaning, the model maintains accuracy requirements while becoming suitable for digital predistortion implementation.
2Measurement precision
If a full physical model is used, then the modeling accuracy is high, but the device complexity increases and cannot be directly applied to DPD technology
Solution Approach 1:
The patent extracts only the essential physical characteristics (thermal effects and trapping effects) from the full physical model that are most relevant to DPD applications. By selectively extracting key features rather than using the complete physical model, the approach maintains necessary accuracy while reducing complexity to DPD-appropriate levels.
Solution Approach 2:
The patent creates a simplified copy of the full physical model that preserves the essential input-output relationships and physical characteristics needed for DPD. This copied model structure maintains accuracy for predistortion purposes while being mathematically tractable and implementable in digital systems.
3Reliability
If circuit level simulation model is used, then physical prior knowledge is incorporated, but the model cannot be directly applied to DPD due to bandpass vs baseband, continuous vs discrete, and numerical vs analytical expression mismatches
Solution Approach 1:
The patent systematically transforms the model across multiple parameter dimensions: frequency domain (bandpass to baseband), time domain (continuous to discrete), and mathematical form (numerical to analytical). Each transformation preserves the essential physical relationships while adapting the model for DPD compatibility.
Solution Approach 2:
Instead of trying to adapt DPD technology to match circuit simulation models, the patent inverts the approach by transforming the circuit simulation model into the format required by DPD. This inversion strategy makes the physical model compatible with DPD requirements while preserving physical accuracy.
Data Source
AI summary
Disclosed is a method and system for forward synthesis of a digital predistortion nonlinear model derived from a circuit description, used for nonlinear modeling of a power amplifier. The method includes establishing a frequency band-continuous time-domain expression, transforming the frequency band expression to a baseband expression, and transforming the baseband-continuous time-domain expression to a baseband-discrete time-domain expression. The present disclosure can obtain the DPD nonlinear model by forward synthesis, provide an accurate model structure and initial parameter values for establishing the RF power amplifier model in the power amplifier nonlinear predistortion technology, make full use of the prior knowledge of the RF power amplifier system to obtain the nonlinear model structure that best conforms to the physical essence, and directly obtain the minimum memory depth and nonlinear order, and can be integrated into an RF power amplifier circuit simulation software.