Sparse Tap Coefficient Vector for Optical Signal Equalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In digital coherent optical transmission, reducing the power consumption of adaptive equalizers in signal processing apparatuses leads to a decrease in the performance of polarization mode dispersion compensation, as the number of taps in adaptive filters must be minimized to conserve power, but this compromises the ability to maintain effective compensation.
Innovation Solution
A signal processing apparatus using an L0 norm-constrained steepest descent method to approximate the transmission line's characteristic with a sparse tap coefficient vector, where tap coefficients below a threshold are set to zero, reducing power consumption while maintaining compensation performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If the number of taps of the adaptive filter is reduced to lower power consumption, then power consumption of the adaptive equalizer is reduced, but the performance of polarization mode dispersion compensation is degraded
Solution Approach 1:
The patent applies L0 norm constraint to the tap coefficient vector, which fundamentally changes the optimization parameter from traditional L2 norm to L0 norm. This parameter change enables sparse solution where most coefficients become zero, allowing reduction in number of active taps while maintaining compensation performance through the constraint that limits the number of non-zero coefficients to a predetermined value or less
Solution Approach 2:
The zeroing unit extracts and removes insignificant coefficients by setting coefficients with absolute values below a threshold to zero. This extraction process separates the essential coefficients that maintain polarization mode dispersion compensation performance from the redundant ones, enabling the system to operate with fewer active taps and thus lower power consumption
2Device complexity
If the number of taps of the adaptive filter is reduced to lower power consumption, then device complexity is reduced, but the performance of signal processing is degraded
Solution Approach 1:
By changing the optimization criterion to L0 norm constraint, the system transforms the complexity-performance tradeoff. The L0 norm constraint creates a sparse coefficient distribution where only a limited number of taps remain active, reducing device complexity while the constraint ensures that these remaining taps maintain the necessary signal processing performance
Solution Approach 2:
The system extracts and eliminates redundant taps through the zeroing operation. By identifying and zeroing coefficients below the threshold, the system removes unnecessary computational elements, reducing device complexity while preserving the essential signal processing functionality through the remaining non-zero coefficients
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A signal processing apparatus includes: a coefficient update unit configured to approximate a characteristic of a transmission line of an optical signal by a first tap coefficient vector of which an L0 norm is a predetermined value or less; a zeroing unit configured to generate a second tap coefficient vector by replacing, with 0, a tap coefficient of which an absolute value is less than a threshold among tap coefficients of the first tap coefficient vector; and an adaptive filter configured to perform, based on the second tap coefficient vector, adaptive equalization processing on a digital signal corresponding to an optical signal received via the transmission line (Figure 2).