Blind Polarisation Demultiplexing With Feedback FIR Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fiber optic systems face challenges in blind polarisation demultiplexing, particularly in fiber links with time-varying polarisation changes, where existing solutions are either complex, expensive, or result in information loss due to polarisation-dependent attenuation and the lack of training sequences in fiber network standards.
Innovation Solution
A method and arrangement for blind polarisation demultiplexing in a coherent receiver using a multidimensional filter that calculates error correction factors and updates filter coefficients based on correlation between x-polarisation and y-polarisation samples, enabling effective separation of orthogonal polarised signals without requiring polarisation controllers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard equalisation algorithms (LMS, CMA) are used for blind polarisation demultiplexing, then initial acquisition and carrier phase compensation are achieved, but degenerative solutions occur where one polarized signal is demultiplexed to both output polarisations causing information loss
Solution Approach 1:
The patent introduces a feedback mechanism that monitors the correlation between x- and y-polarisation outputs and uses this information to dynamically adjust filter coefficients. The error correction factor is calculated based on the correlation between output signals and fed back to update the multidimensional filter, preventing degenerative solutions where information is lost to both outputs
Solution Approach 2:
The patent introduces an intermediary error correction mechanism that mediates between the two polarisation outputs. By calculating error correction factors based on correlation and using these to adjust filter coefficients, the system prevents direct conflict between x and y polarisation signals that would cause information loss
2Adaptability or versatility
If adaptive optical polarisation controllers are used to handle time-varying polarisation changes, then polarisation alignment is achieved, but device complexity and cost increase significantly
Solution Approach 1:
The patent replaces mechanical/optical polarisation controllers with a digital signal processing solution. Instead of using physical components to adjust polarisation, the invention uses software-based algorithms that process the electrical domain signals from the coherent receiver, achieving polarisation tracking without mechanical or optical moving parts
Solution Approach 2:
The system performs self-adjustment through automatic calculation of error correction factors and dynamic update of filter coefficients based on real-time correlation analysis. The algorithm autonomously tracks and compensates for polarisation changes without external control inputs or manual intervention
3Reliability
If polarisation controllers are used for fibre links with time-varying polarisation, then signal separation is improved, but the system becomes complicated and expensive
Solution Approach 1:
The patent replaces complex optical polarisation control hardware with digital signal processing algorithms. The coherent receiver converts optical signals to electrical domain, where polarisation diversity is handled through software-based multidimensional filtering and error correction, eliminating the need for complex optical controllers
Solution Approach 2:
The multidimensional filter serves multiple functions simultaneously: it performs polarisation demultiplexing, compensates for polarisation-dependent loss, and adapts to time-varying polarisation changes. This single computational structure replaces what would otherwise require multiple separate optical components and controllers
Data Source
AI summary
A received optical signal is coherently demodulated and converted into orthogonal x-polarization samples, and y-polarization samples. These samples are converted into signal x-samples and signal y-samples by an FIR butterfly filter. Correction values are calculated in an error calculating circuit of a control unit and added to filter transfer functions derived by a standard algorithm to determine corrected filter coefficients. Degenerate convergences calculating the transfer functions are avoided.