DPD Estimation Window Search for Transmitter Linearity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional digital pre-distortion (DPD) signal processing systems face significant variance in linearity performance due to the randomness of live transmitted waveforms, leading to increased latency, memory requirements, and complexity, as well as inefficient degradation detection and averaging of DPD coefficients.

Innovation Solution

The proposed solution involves a DPD estimation window search technique that analyzes signal power characteristics to determine an optimal estimation window, reducing the number of required signal samples and improving linearity and efficiency by selecting samples based on detected peaks and power thresholds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional DPD estimation uses a large number of signal samples to reduce variance, then linearity performance consistency improves, but latency and memory requirements increase

Engineering Contradiction:
Improvelinearity performance consistencyVSAvoidlatency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts only the most informative signal samples (those containing peaks above a threshold) from the complete signal set for DPD estimation. By selecting a subset of high-value samples rather than using all available samples, the system achieves consistent linearity performance with reduced memory requirements and lower latency, as it processes fewer samples while maintaining estimation quality.

Inventive Principle:
Principle #2Taking out (Extraction)

2Reliability

If conventional DPD estimation uses a large number of signal samples to reduce variance, then linearity performance consistency improves, but device complexity increases

Engineering Contradiction:
Improvelinearity performance consistencyVSAvoidcomplexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts only the most informative signal samples (those containing peaks above a threshold) from the complete signal set for DPD estimation. By selecting a subset of high-value samples rather than using all available samples, the system achieves consistent linearity performance with reduced memory requirements and lower latency, as it processes fewer samples while maintaining estimation quality.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies different processing quality to different portions of the signal by identifying and selectively processing only those segments containing peaks above a threshold. This local quality approach focuses computational resources on the most informative portions of the signal, reducing overall system complexity while maintaining estimation accuracy.

Inventive Principle:
Principle #3Local quality

3Reliability

If conventional DPD systems average coefficients from multiple estimations to reduce variance, then performance consistency improves, but productivity decreases

Engineering Contradiction:
Improveperformance consistencyVSAvoidcoefficient update efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent performs preliminary selection of high-quality samples containing peaks before conducting DPD estimation. By pre-filtering the signal to identify and select only the most informative samples, the system achieves consistent performance in a single estimation pass, eliminating the need for multiple estimations and coefficient averaging, thus improving productivity while maintaining reliability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4016947A1Digital pre-distortion (DPD) estimation window search
Publication Date: 2022.06.22 INTEL CORP
  • EP4016947A1 patent drawingFigure 1
  • EP4016947A1 patent drawingFigure 2
  • EP4016947A1 patent drawingFigure 3~4

AI summary

An apparatus for transmitting output signals includes a pre-distortion circuit configured to perform digital pre-distortion (DPD) on input signal samples using a plurality of DPD coefficients to generate a plurality of pre-distorted samples. The apparatus further includes an estimation window search circuit configured to determine a signal power characteristic for each pre-distorted sample of the plurality of pre-distorted samples, and select a subset of the plurality of pre-distorted samples based on the determined signal power characteristics. The subset of the plurality of pre-distorted samples is selected to fit within a predetermined estimation window size. The plurality of DPD coefficients is updated based on the subset of the plurality of pre-distorted samples.