Power-Feature ML DPD Modeling for Nonlinear Distortion Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing approaches for non-linear device and/or digital predistortion (DPD) behavior modeling struggle to accurately model complex non-linearities and memory effects, especially under dynamic traffic conditions and with long-term memory effects.
Innovation Solution
A computer-implemented method using a power feature aided machine learning (ML) model that extracts and labels power features from input signals to model the behavior of DPD, reducing non-linear distortion of output signals from non-linear devices. The method includes training and applying a tree-based power feature aided gradient boosting or extreme gradient boosting model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing DPD modeling approaches (MP/GMP) are used, then the device complexity is low, but the modeling precision deteriorates under dynamic traffic conditions and complex non-linearities
Solution Approach 1:
The patent transforms the DPD modeling approach by changing the parameter representation from traditional polynomial coefficients to machine learning model parameters (weights and biases). This allows the system to capture complex non-linearities and memory effects more accurately while adapting to dynamic traffic conditions, resolving the contradiction between modeling precision and device complexity
Solution Approach 2:
The patent replaces the mechanical/mathematical polynomial-based DPD system with a machine learning-based system. This substitution enables the model to learn complex patterns from data rather than relying on predefined mathematical relationships, significantly improving modeling precision for dynamic conditions while the computational complexity remains manageable through efficient ML algorithms
2Adaptability or versatility
If traditional DPD models are used, then the computational complexity is low, but the adaptability to dynamic traffic conditions deteriorates
Solution Approach 1:
The patent introduces dynamics into the DPD model by using machine learning algorithms that can adapt to changing traffic conditions. The ML model continuously learns from incoming data and adjusts its parameters accordingly, enabling the system to handle dynamic traffic effects, memory effects, and non-linearities that traditional static models cannot capture
Solution Approach 2:
The patent performs preliminary training of the machine learning model using training data that represents various traffic conditions. This preliminary action prepares the model to handle dynamic traffic scenarios effectively when deployed, allowing it to generalize from training data to unseen dynamic conditions without requiring complex real-time computations
Data Source
AI summary
A computer-implemented method performed by a device configured with a power feature aided machine learning, ML, model is provided that models a behavior of a DPD to reduce non-linear distortion of an output signal of a non-linear device. The method includes extracting a plurality of power features from an input signal destined to be input to the DPD. The method further includes labelling the extracted plurality of power features to obtain at least one labelled average power level; inputting the at least one labelled average power level to the input of the ML model to obtain an output signal from the ML model having characteristics to reduce the non-linear distortion of the output signal of the non-linear device; and providing the output signal from the ML model as an input to the non-linear device.