Capacity-Based Digital Pre-Distortion Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The Minimum Mean Square Error (MMSE) solution for digital pre-distortion (DPD) optimization is not always optimal for communication systems, particularly in terms of channel capacity, and existing methods rely on inaccurate assumptions such as the replaceability of pre-distortion and post-distortion coefficient sets.

Innovation Solution

A capacity-based DPD optimization method that involves choosing a DPD adaptation stimulus for SNR estimation, initializing DPD coefficients, calculating a cost function based on estimated SNR, and numerically optimizing these coefficients to update the DPD set, using techniques like gradient search or pattern search algorithms.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If MMSE solution is used for DPD optimization, then implementation simplicity is improved, but channel capacity optimization deteriorates

Engineering Contradiction:

Improveimplementation simplicityVSAvoidchannel capacity optimization

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The patent changes the optimization parameter from MMSE (minimum mean square error) to capacity-based optimization. By using capacity as the optimization criterion instead of MMSE, the system achieves better channel capacity performance while maintaining a practical implementation through iterative optimization algorithms.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent implements feedback mechanisms where the system measures actual channel capacity performance and uses this information to iteratively adjust DPD coefficients. This feedback loop allows the system to converge to capacity-optimal solutions rather than relying on suboptimal MMSE assumptions.

Inventive Principle:

Principle #23Feedback

2Device complexity

If backward prediction solution is assumed identical to MMSE solution, then device complexity is reduced, but measurement precision deteriorates

Engineering Contradiction:

Improvealgorithm complexityVSAvoidSNR estimation accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent inverts the conventional approach by not assuming backward prediction equals MMSE solution. Instead, it explicitly models the relationship between backward prediction and actual capacity optimization, using the former as a starting point for iterative refinement toward the latter, thereby improving SNR estimation accuracy without excessive complexity.

Inventive Principle:

Principle #13The other way round (Inversion)

3Measurement precision

If iterative MMSE variants are used to improve estimation, then measurement precision is improved, but use of energy increases

Engineering Contradiction:

Improveestimation accuracyVSAvoidpower consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies partial iteration - using iterative optimization but stopping when capacity convergence is achieved rather than pursuing full MMSE convergence. This partial action approach achieves sufficient estimation accuracy for capacity optimization while consuming less energy than full iterative MMSE solutions.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12057870B2Systems and methods for capacity-based digital pre-distortion optimization

Publication Date: 2024.08.06 SEQUANS COMMUNICATIONS

AI summary

Some embodiments relate to systems and methods for capacity based DPD optimization. An example method of capacity based DPD optimization includes choosing a DPD adaptation stimulus to facilitate a SNR estimation. The example method also includes initializing DPD coefficients based on setting the DPD coefficients to an initial guess. Additionally, the example method includes estimating the SNR and calculating the cost function based on the estimated SNR; The example method also includes numerically optimizing the DPD coefficients and updating the DPD coefficient set based on the numerical optimization.