Power Amplifier Predistortion by Power-Interval Coefficient Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing predistortion technologies for power amplifiers suffer from high algorithm complexity and low accuracy due to large predistortion model orders, leading to deteriorated performance in digital and analog predistortion processing.
Innovation Solution
The method involves dividing the non-linear curve of a power amplifier into multiple intervals based on input power magnitude, determining predistortion coefficients at a granular level for each interval, and performing predistortion processing accordingly, thereby reducing algorithm complexity and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If predistortion processing is performed on an entire segment of a non-linear curve of the PA, then the predistortion model order is large, but this results in high algorithm complexity and low accuracy of the predistortion coefficient
Solution Approach 1:
The non-linear curve of the power amplifier is divided into multiple intervals based on input power magnitude. Instead of processing the entire curve segment as one unit, the patent segments the curve into distinct intervals and determines predistortion coefficients for each interval separately. This segmentation reduces the predistortion model order for each interval, thereby lowering algorithm complexity while improving the accuracy of predistortion coefficients through more granular processing.
2Reliability
If predistortion processing is performed on an entire segment of a non-linear curve of the PA, then the coverage is complete, but this results in deterioration of predistortion performance
Solution Approach 1:
The patent applies local quality by determining predistortion coefficients at a granular level for each interval of the non-linear curve rather than using a single model for the entire segment. Each interval can have its own optimized predistortion coefficients tailored to the specific characteristics of that power range, improving overall predistortion performance while keeping the model order manageable for each local segment.
Data Source
AI summary
In one example method, a first device obtains configuration information of a reference signal, an interval range of each of a plurality of intervals, and a predistortion model parameter of each of the intervals, where the plurality of intervals are power intervals of the reference signal. The first device obtains a first reference signal based on the configuration information. The first device receives a reference signal from a second device to obtain a second reference signal. The first device determines, based on a predistortion model parameter of a first interval in the plurality of intervals, the first reference signal, and the second reference signal, a predistortion coefficient of the first interval.