PA Predistorter Space Mapping for High-Fidelity Linearization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing power amplifier linearization techniques, such as polynomial and artificial neural network models, fail to provide high fidelity compensation for non-linearity, leading to inefficiencies in power amplifier operation.
Innovation Solution
The use of space mapping based predistortion, which involves a coarse model and a fine model of the power amplifier's non-linear characteristic, with an adaptation sub-system that configures and trains these models to achieve high fidelity compensation, effectively counteracting the power amplifier's non-linear behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If polynomial or artificial neural network models are used for predistortion, then the power amplifier non-linearity can be compensated, but the modeling fidelity is insufficient and computational complexity increases
Solution Approach 1:
The predistorter is segmented into two independent models: a memoryless predistortion model that handles amplitude-to-amplitude and amplitude-to-phase distortions, and a separate memory effect model that handles temporal dependencies. This segmentation allows each model to be optimized independently, achieving high modeling fidelity without excessive computational complexity.
Solution Approach 2:
The system dynamically adapts the memory depth parameter based on signal conditions and training progress. The memory depth is not fixed but can be adjusted during operation to balance between modeling accuracy and computational load, allowing the system to optimize performance for different operating scenarios.
2Measurement precision
If the predistorter model is trained adaptively, then the compensation accuracy improves, but the training time and convergence speed become problematic
Solution Approach 1:
The system performs preliminary initialization of predistorter coefficients using analytical solutions or pre-trained values before entering the adaptive training phase. This preliminary action provides a good starting point for the optimization algorithm, significantly reducing the number of iterations needed to converge and thus reducing overall training time.
Solution Approach 2:
The system implements a feedback mechanism where the output of the power amplifier is measured and fed back to the adaptation subsystem. This feedback is used to continuously update the predistorter coefficients, allowing the system to achieve high compensation accuracy through iterative refinement while monitoring convergence to stop training early when sufficient accuracy is reached.
3Loss of energy
If the power amplifier operates at maximum efficiency, then power dissipation is reduced, but linearity deteriorates due to non-linear operation
Solution Approach 1:
The predistorter applies preliminary anti-action by pre-distorting the input signal in the opposite direction of the power amplifier's non-linearity. This anticipatory correction ensures that when the signal passes through the non-linear power amplifier, the distortions cancel out, achieving linear output while maintaining high efficiency operation.
Solution Approach 2:
The predistorter acts as an intermediary between the input signal and the power amplifier. It modifies the signal characteristics before amplification, creating a pre-distorted signal that, when amplified by the non-linear PA, produces a linear output. This intermediary function allows the PA to operate in its high-efficiency non-linear region while maintaining signal linearity.
Data Source
AI summary
Systems and methods are disclosed for compensating for non-linearity of a power amplifier using space mapping based predistortion. In one embodiment, a transmitter includes a power amplifier that amplifies a power amplifier input signal, a predistorter that effects predistortion of the power amplifier input signal to compensate for a non-linear characteristic of the power amplifier using a space mapping based model of an inverse of the non-linear characteristic of the power amplifier, and an adaptation sub-system that adaptively configures the space mapping based model of the non-linear characteristic of the power amplifier. In one embodiment, the adaptation sub-system adaptively configures a space mapping based model of the non-linear characteristic of the power amplifier and adaptively configures the space mapping based model of the inverse of the non-linear characteristic of the power amplifier based on the space mapping based model of the non-linear characteristic of the power amplifier.