Baseband Predistortion with Piecewise LUT Memory Compensation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing baseband predistortion methods struggle to effectively compensate for both RF frequency response memory effects and bias-induced or thermal memory effects in multi-channel wideband wireless transmitters, leading to inefficiencies and computational complexity.

Innovation Solution

A piecewise pre-equalized lookup table (LUT) based predistortion system that reduces computational complexity and numerical instability, capable of compensating for electrical and thermal memory effects while maintaining linearity performance comparable to memory polynomial PD algorithms.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If memory polynomial PD algorithm is used to compensate memory effects, then linearity performance is improved, but computational complexity increases significantly

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

Solution Approach 1:

The predistorter is divided into two independent functional blocks: a memoryless nonlinearity compensator and a memory effects compensator. This segmentation allows each block to handle specific distortion types separately, reducing the overall computational complexity while maintaining linearity performance comparable to memory polynomial methods.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The memory effects compensation function is extracted from the complex memory polynomial algorithm and implemented as a separate FIR filter block. This extraction simplifies the computational structure by isolating the memory compensation task from the nonlinearity compensation, reducing the overall computational burden.

Inventive Principle:
Principle #2Taking out (Extraction)

2Reliability

If Volterra filter structure is used to compensate memory effects, then memory effect compensation is achieved, but the number of optimization coefficients becomes very large

Engineering Contradiction:
Improvememory effect compensationVSAvoidnumber of coefficients
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The memory effects compensation is extracted as a separate FIR filter block with significantly fewer coefficients compared to the full Volterra filter structure. This extraction reduces the number of optimization coefficients while maintaining memory effect compensation capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The predistorter is segmented into distinct functional blocks, with memory effects handling separated from nonlinearity compensation. This segmentation reduces the total number of coefficients that need to be optimized compared to the monolithic Volterra approach.

Inventive Principle:
Principle #1Segmentation

3Device complexity

If Hammerstein predistorter is used to reduce complexity, then computational complexity is reduced, but performance deteriorates because it cannot correct all types of memory effects

Engineering Contradiction:
Improvecomputational complexityVSAvoidmemory effect compensation performance
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The invention merges the strengths of both memoryless and Hammerstein approaches by combining a memoryless nonlinearity compensator with a separate memory effects compensator (FIR filter). This combination achieves comprehensive memory effect compensation while maintaining computational simplicity, overcoming the limitations of the Hammerstein predistorter.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3416340B1Method and system for baseband predistortion linearization in multi-channel wideband communication systems
Publication Date: 2020.10.21 DALI SYST LTD
  • EP3416340B1 patent drawingFigure 1
  • EP3416340B1 patent drawingFigure 2
  • EP3416340B1 patent drawingFigure 3A~3E

AI summary

An efficient baseband predistortion linearization method for reducing the spectral regrowth and compensating memory effects in wideband communication systems using effective multiplexing modulation technique such as wideband code division multiple access and orthogonal frequency division multiplexing is disclosed. The present invention is based on the method of piecewise pre-equalized lookup table based predistortion, which is a cascade of a lookup table predistortion and piecewise pre- equalizers, to reduce the computational complexity and numerical instability for desired linearity performance with memory effects compensation for wideband transmitter systems. Therefore, the present invention could reduce the computational load, which saves hardware resources in an implementation and improve performance, in terms of adjacent channel power ratio.