Dual Sensor Receiver for Self-FEXT Mitigation in Vectored xDSL Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital subscriber line (xDSL) systems face significant challenges in mitigating crosstalk noise, particularly self-induced far-end crosstalk (self-FEXT) and external noise sources like radio frequency interference (RFI) and power line communications (PLC), which affect the signal-to-noise ratio (SNR) and system performance.
Innovation Solution
A dual sensor system is implemented in a vectored environment, where a second sensor learns coefficients to remove self-FEXT and external noise, and a linear combiner is applied to combine information from both sensors, effectively mitigating residual self-FEXT and improving SNR by compensating for alien disturbers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single sensor is used in a vectored system, then self-FEXT mitigation can be achieved through coefficient learning, but external noise sources such as RFI, PLC, and common mode noise cannot be effectively addressed
Solution Approach 1:
The patent combines information from multiple sensors (main sensor and second sensor) to create a multi-sensor receiver that can simultaneously mitigate self-FEXT and address external noise sources. The linear combiner merges signals from both sensors, allowing the system to leverage the strengths of each sensor for comprehensive noise cancellation.
Solution Approach 2:
The second sensor is designed to detect multiple types of disturbances including self-FEXT, external noise, RFI, PLC, and common mode noise. This universal sensing capability allows a single system to handle diverse noise sources, making the receiver adaptable to various interference conditions without requiring separate specialized systems.
2Adaptability or versatility
If multiple sensors are deployed to address diverse noise sources, then system versatility and noise mitigation capability improve, but system complexity and computational requirements increase
Solution Approach 1:
The patent segments the noise mitigation function into distinct modules: self-FEXT mitigation coefficients learned by the vectored system, second sensor coefficients learned at the second sensor, and a linear combiner that combines both information. This modular segmentation manages complexity by organizing the multi-sensor system into manageable, independently learnable components.
Solution Approach 2:
The system performs preliminary self-FEXT mitigation using the main sensor and vectored system before processing signals through the linear combiner. This preliminary action reduces the computational burden on subsequent stages by pre-processing the dominant self-FEXT component, allowing the second sensor and combiner to focus on residual noise and external disturbances.
3Measurement precision
If self-FEXT mitigation coefficients are applied to remove self-FEXT, then signal quality improves, but residual self-FEXT remains that requires further mitigation
Solution Approach 1:
The patent implements a feedback mechanism where the linear combiner uses information from the second sensor to identify and cancel residual self-FEXT that remains after initial mitigation. The system continuously monitors the victim line and adjusts the linear combiner coefficients to eliminate remaining self-FEXT components, ensuring high cancellation accuracy through iterative refinement.
Solution Approach 2:
The linear combiner acts as an intermediary between the main sensor and the final received signal. It processes the output from the main sensor and incorporates information from the second sensor to eliminate residual self-FEXT, serving as a mediating stage that refines the signal before it reaches the final decision device.
Data Source
AI summary
In accordance with one embodiment, a method is implemented in a vectored system for improving a signal-to-noise ratio (SNR) of a far end transmitted signal on a victim line in the system. The method comprises mitigating, by the vectored system, self-induced far-end crosstalk (self-FEXT) on the victim line based on self-FEXT mitigation coefficients and receiving, by a second sensor, information relating to at least one of: self-FEXT of the vectored system, external noise, and the far end transmitted signal. The method further comprises learning, at the second sensor, coefficients relating to self-FEXT coupling into the second sensor and removing self-FEXT from the second sensor based on the learned coefficients. Upon removal of self- FEXT from the second sensor, a linear combiner configured to combine information relating to the victim line and the second line is learned.