Fractional-Sampling FIR Equalization for Optical Signal Distortion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing fractional sampling methods in digital coherent optical transmission face challenges with slow response speed and increased power consumption due to limited updateable tap coefficient sequences and frequency of coefficient updates.

Innovation Solution

A signal processing device and method utilizing an FIR filter with non-integer sampling frequencies and adaptive equalization processing, where filter coefficients are updated using gradient vectors and shift matrices to classify symbols into sequences, enabling high-speed control while reducing power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If fractional sampling is used to reduce power consumption and resampling processing, then power consumption decreases and baud rate increases, but the updatable tap coefficient sequence changes and response speed to disturbance decreases

Engineering Contradiction:
Improvepower consumptionVSAvoidresponse speed
Core Design Contradiction:
Use of energy by moving objectVSSpeed

Solution Approach 1:

The patent divides the tap coefficient updates into multiple sequences (first sequence, second sequence, etc.) where different coefficients are updated at different times. This segmentation allows the system to maintain lower power consumption by updating coefficients less frequently while still providing periodic updates across multiple sequences, thus balancing power savings with response speed to disturbances.

Inventive Principle:
Principle #1Segmentation

2Use of energy by moving object

If the number of taps of FIR filter is reduced to decrease power consumption, then power consumption decreases, but polarization mode dispersion compensation capability is reduced

Engineering Contradiction:
Improvepower consumptionVSAvoidpolarization mode dispersion compensation capability
Core Design Contradiction:
Use of energy by moving objectVSReliability

Solution Approach 1:

The patent employs multiple tap coefficient sequences that can be dynamically selected and updated. By maintaining multiple sequences with different update timings rather than a single static set of coefficients, the system achieves better polarization mode dispersion compensation with fewer taps per sequence, as the combined effect of multiple sequences provides the necessary compensation capability while keeping individual sequence complexity low.

Inventive Principle:
Principle #15Dynamics

3Ease of operation

If 2×2 complex filter arrangement is employed for polarization separation, then polarization separation is achieved, but power consumption per tap becomes 16 times larger than 1×1 real-number filter

Engineering Contradiction:
Improvepolarization separation capabilityVSAvoidpower consumption per tap
Core Design Contradiction:
Ease of operationVSUse of energy by moving object

Solution Approach 1:

The patent changes the operational parameters of the FIR filter by using multiple tap coefficient sequences with different update timings and classifications. This parameter change allows the system to achieve polarization separation with reduced computational complexity per tap, as the sequences are updated selectively rather than requiring continuous full-complexity processing, thereby reducing power consumption while maintaining polarization separation capability.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4693954A1Signal processing device and signal processing method
Publication Date: 2026.02.11 NT T INC
  • EP4693954A1 patent drawingFigure 1
  • EP4693954A1 patent drawingFigure 2
  • EP4693954A1 patent drawingFigure 3

AI summary

Provided is a signal processing device including: a finite impulse response (FIR) filter in which a sampling frequency of an input signal input from an outside is at a ratio of less than two samples per symbol or at a ratio which is not capable of being represented by an integer, and the sampling frequency of an output signal is one sample per symbol, the FIR filter classifying each symbol into any of a plurality of sequences according to a positional relationship of sampling points of the input signal with respect to the output signal, and executing adaptive equalization processing on a digital signal corresponding to an optical signal received via an optical transmission line on the basis of filter coefficients different for each sequence; and a coefficient update unit that updates the filter coefficients of a reference sequence by using a gradient vector obtained from a known signal or a data signal to set the updated filter coefficients in the FIR filter, and generates filter coefficients of a sequence other than the reference sequence by performing an operation on a square matrix different for each sequence for the filter coefficients of the reference sequence to set the generated filter coefficients in the FIR filter.