Digital Predistortion with Embedded Upsampling for Transmitter Chains

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing digital signal predistortion methods face limitations in achieving high performance while maintaining low complexity, particularly due to the high complexity and cost associated with processing non-linear characteristics of power amplifiers and in-phase and quadrature modulators in communication systems.

Innovation Solution

The method involves decomposing input signals into subsets and applying a non-linear model to each subset, followed by upsampling within the predistortion process using a polyphase structure, allowing for improved performance with reduced complexity by embedding upsampling into the predistortion process itself.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If digital predistortion processing is applied to mitigate non-linear effects of power amplifiers and IQM, then system performance is improved, but computational complexity increases

Engineering Contradiction:
Improvesystem performanceVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the input signal into multiple subsets and processes each subset separately through parallel processing paths. This segmentation allows the system to achieve high-performance predistortion by processing signal subsets in parallel, thereby improving computational efficiency while maintaining low complexity per processing unit.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent embeds upsampling operation within the predistortion processing itself, transitioning from traditional sequential processing (predistortion then upsampling) to a multi-dimensional approach where upsampling factors are applied during the predistortion stage. This dimensional change in processing architecture enables the system to achieve high performance with reduced overall complexity.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If upsampling is performed after digital predistortion processing, then higher performance is achieved, but implementation complexity and cost increase

Engineering Contradiction:
ImproveperformanceVSAvoidimplementation complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent merges the upsampling operation with the predistortion processing into a unified computational framework. By combining these two separate operations into one integrated process, the system eliminates the need for separate upsampling hardware or software modules, thereby reducing implementation complexity and cost while achieving the same performance benefits.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent performs upsampling as a preliminary action during the predistortion processing stage rather than as a subsequent step. This preliminary upsampling allows the system to prepare the signal at the appropriate resolution before predistortion computation, simplifying the overall architecture and reducing the complexity of later processing stages.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4278482B1Method and apparatus for predistortion of a digital signal in a transmitter chain
Publication Date: 2025.01.01 HUAWEI TECH CO LTD
  • EP4278482B1 patent drawingFigure 1A
  • EP4278482B1 patent drawingFigure 1B
  • EP4278482B1 patent drawingFigure 2A

AI summary

An apparatus (100) for predistortion of a digital signal in a transmitter chain comprises a feeder (102), a decomposer (104), an evaluator (106), a predistortion unit (108) and a composer (110). An input signal (202) sampled at a first sampling rate is obtained by the feeder (102) and is decomposed into k_down input signal subsets of equal length L by the decomposer (104). K_down model input signal subsets of equal length L are generated by the evaluator (106) by means of applying a non-linear model to every input signal subset. The k_down model input signal subsets are processed by the predistortion unit (108) to obtain k_up predistortion signal subsets of equal length L which then are merged by the composer (110) into a predistortion signal with a second sampling rate which is k_up/k_down times the first sampling rate.