Memory Polynomial Predistortion for Wideband Power Amplifier Nonlinearity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

High power amplifiers in satellite communications systems cause significant non-linear and memory effect distortions, leading to reduced throughput and performance, which existing pre-distortion techniques, such as symbol and sample pre-distortion, are unable to fully address effectively.

Innovation Solution

A multi-rate, iterative memory polynomial model is employed for pre-distorting digital signals before amplification to correct for these distortions, using a pre-distorter that models the high power amplifier and associated components, optimizing pre-distorter taps through iterative and gradient-based methods to minimize error vector magnitude.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If high power amplifiers operate at or near saturation level to provide high throughput and increased efficiency, then power efficiency and throughput are improved, but significant non-linear distortions are generated that degrade signal quality and channel performance

Engineering Contradiction:
ImprovethroughputVSAvoidnon-linear distortions
Core Design Contradiction:
ProductivityVSObject-generated harmful factors

Solution Approach 1:

The patent applies pre-distortion to the digital signal before amplification, which is a preliminary action that anticipates and compensates for the non-linear distortions that will occur during amplification. By pre-shaping the signal in the digital domain to counteract the expected amplifier non-linearities, the system maintains high throughput operation while reducing signal degradation at the amplifier output.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent employs a multi-rate memory polynomial model that dynamically adjusts pre-distortion parameters based on the amplifier's operating conditions. By changing parameters such as the pre-distortion coefficients and sampling rates according to the amplifier state, the system adapts to varying non-linear characteristics while maintaining optimal performance across different throughput levels.

Inventive Principle:
Principle #35Parameter changes

2Power

If high power amplifiers operate at saturation to maximize power output, then radiated power is increased, but heat generation increases temperature of the amplifier and surrounding electronics, degrading their performance

Engineering Contradiction:
Improveradiated powerVSAvoidamplifier temperature
Core Design Contradiction:
PowerVSTemperature

Solution Approach 1:

By applying pre-distortion in the digital domain before the signal reaches the power amplifier, the system compensates for non-linear effects that would otherwise require higher power operation to overcome. This allows the amplifier to operate at optimal power levels without excessive heat generation, as the signal quality is maintained through digital compensation rather than brute-force power increases.

Inventive Principle:
Principle #10Preliminary action

3Object-generated harmful factors

If existing pre-distortion techniques such as symbol pre-distortion are used, then some distortion correction is achieved, but performance and throughput are limited

Engineering Contradiction:
Improvedistortion correctionVSAvoidthroughput
Core Design Contradiction:
Object-generated harmful factorsVSProductivity

Solution Approach 1:

The patent replaces traditional symbol-level pre-distortion methods with a multi-rate memory polynomial model that operates at higher sampling rates. This substitution enables more precise correction of non-linear and memory effects by processing individual signal samples rather than just symbol boundaries, thereby achieving both better distortion correction and higher throughput capability.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent implements a dynamic pre-distortion system that adapts to changing amplifier characteristics and operating conditions. By using a memory polynomial model that accounts for time-varying non-linearities and operating at multiple rates, the system dynamically adjusts pre-distortion parameters to maintain optimal performance across varying throughput levels, unlike static symbol-level methods.

Inventive Principle:
Principle #15Dynamics

4Object-generated harmful factors

If sample pre-distortion is used to correct wider bandwidth and more prevalent non-linear and memory distortions, then distortion correction capability is improved, but computational complexity and processing requirements increase

Engineering Contradiction:
Improvenon-linear and memory distortionsVSAvoidprocessing complexity
Core Design Contradiction:
Object-generated harmful factorsVSDevice complexity

Solution Approach 1:

The patent segments the pre-distortion processing into multiple operational modes or rates, allowing the system to apply appropriate levels of complexity based on operating conditions. The multi-rate memory polynomial model can operate at different sampling rates and polynomial orders, enabling the system to balance distortion correction capability with computational complexity by selecting the appropriate processing level for each situation.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3776848B1Multirate, iterative, memory polynomial based modeling and pre-distortion of high bandwidth power amplifiers
Publication Date: 2024.01.24 NORTHROP GRUMMAN SYSTEMS CORP
  • EP3776848B1 patent drawingFigure 1~2
  • EP3776848B1 patent drawingFigure 3~5
  • EP3776848B1 patent drawing

AI summary

A system and method for pre-distorting a digital signal in a digital communications system. The method includes converting digital bits to be transmitted to a series of symbols defining the bits and providing the symbols to a pulse shaping filter (PSF) that provides samples of the symbols at a predetermined sample rate. The method also includes providing the filtered samples to a pre-distorter that pre-distorts the samples, wherein pre-distorting the samples includes providing a non-linear transformation of the samples that is defined by pre-distorter taps, and providing the pre-distorted samples to a power amplifier to be transmitted, where pre-distorting the samples includes modeling the power amplifier and associated transmitter components using a degree three memory polynomial.