Adaptive Filter Circuit Reducing Calculation Overhead in Optical Receivers

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In adaptive multi-layer filters used for optical fiber communication, the recalculation of intermediate filter outputs during coefficient updates leads to an increase in calculation amount, making it challenging to compensate for various distortions efficiently.

Innovation Solution

The proposed solution involves an equalization signal processing circuit with a filter group where filters before the final stage convolve their coefficients with both the current and previous input signal vectors, and output these results to subsequent filters, thereby reducing the need for full recalculation of intermediate outputs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If adaptive multi-layer filters are used to compensate for various distortions in optical fiber communication, then distortion compensation performance is improved, but calculation amount increases due to recalculation of intermediate filter outputs during coefficient updates

Engineering Contradiction:
Improvedistortion compensation performanceVSAvoidcalculation efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing intermediate filter outputs before coefficient updates occur. When coefficient updates are performed, the system retrieves previously calculated intermediate outputs instead of recalculating them, thereby reducing the calculation amount while maintaining distortion compensation performance.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses copying by creating a copy of the intermediate filter outputs that were calculated in previous time steps. These copied values are then used during coefficient updates, avoiding the need to recalculate the entire filter chain and reducing computational complexity while preserving compensation effectiveness.

Inventive Principle:
Principle #26Copying

2Measurement precision

If full recalculation of intermediate filter outputs is performed during coefficient updates, then coefficient accuracy is maintained, but processing time increases

Engineering Contradiction:
Improvecoefficient accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary calculations of intermediate filter outputs and stores them before coefficient updates. During updates, these pre-calculated values are retrieved and used, allowing coefficient accuracy to be maintained without requiring full recalculation, thus reducing processing time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent discards the need to recalculate intermediate outputs that have already been computed in previous time steps. By recovering and reusing these previously calculated intermediate values, the system maintains coefficient accuracy while significantly reducing the processing time required during coefficient updates.

Inventive Principle:
Principle #34Discarding and recovering

Data Source

PatentUS20250141562A1Communication system, receiver, equalization signal processing circuit, method, and computer readable medium
Publication Date: 2025.05.01 NEC CORP
  • US20250141562A1 patent drawing
  • US20250141562A1 patent drawing
  • US20250141562A1 patent drawing

AI summary

A receiving circuit coherent-receives a signal being transmitted from a transmitter via a transmission path. An equalization signal processing circuit performs equalization signal processing with respect to the reception signal being coherent-received. A filter group includes a plurality of filters being connected in series along a signal path of the reception signal. A coefficient updating means adaptively controls filter coefficients of at least some filters of the plurality of filters, by using an error back propagation method. Among the plurality of filters, at least some of intermediate filters convolve the filter coefficients with respect to an input signal vector, and output a calculation result of the convolution and a calculation result of the convolution that is acquired by calculation at a previous time, to a filter in a following stage.