DSL Disturber Identification via Noise Signature Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for identifying disturbing power supply units (PSUs) in Digital Subscriber Lines (DSLs) often lead to misdiagnosis, particularly when PSU-induced noise is indirectly induced in DSL lines within a binder, causing incorrect repair advice for customers.
Innovation Solution
A device and method that collect noise data as a function of frequency, detect noise signatures of disturbing PSUs, group impacted lines, and identify the line with the most severe noise signature as the disturber, enabling accurate diagnosis and repair advice by distinguishing directly and indirectly affected lines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise signature detection is applied to identify disturbing PSUs in DSL lines, then the ability to detect PSU-induced noise is improved, but misdiagnosis occurs when noise is indirectly induced in neighboring lines within a binder
Solution Approach 1:
The invention segments the binder into multiple groups based on noise signature correlations. By dividing the lines into distinct groups where each group shares a common disturbing PSU, the system can identify the specific disturber within each group rather than falsely attributing noise to every line. This segmentation resolves the contradiction by maintaining detection sensitivity while improving diagnostic reliability through grouped analysis.
Solution Approach 2:
The invention introduces noise signature correlation as an intermediary mechanism to distinguish direct victims from indirect victims. By calculating correlation values between noise signatures of different lines, the system identifies which lines share the same disturbing source. This intermediary correlation analysis prevents misdiagnosis by mediating between the detected noise presence and the final identification of the disturbing PSU location.
2Ease of operation
If per-line noise analysis is performed to identify disturbing PSUs, then individual line diagnostics are improved, but crosstalk between lines causes false identification of disturbers
Solution Approach 1:
The invention maintains the simplicity of per-line analysis by segmenting lines into correlation-based groups. Each group is analyzed independently to identify its specific disturber, preserving the ease of operation while eliminating false identifications caused by crosstalk. The segmentation allows systematic processing of individual lines within the context of their correlation groups.
Solution Approach 2:
The invention changes the analysis parameter from individual line noise levels to noise signature correlation values. By transforming the data into correlation metrics that compare noise patterns across lines, the system maintains operational simplicity while significantly improving disturber identification accuracy. This parameter transformation distinguishes direct victims (high correlation with disturber) from indirect victims (lower correlation).
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In order to identify a line (d1, d2, d3) holding a disturbing power supply unit for a plurality of Digital Subscriber Lines (d1, v11, v12, v13, v14, v15, d2, v21, v22, v23, d3, v31, v32, v33, v34, ...), noise data as a function of frequency (221, 222, 223, 224, 225, 226; 311, 312, 313, ... 31 M) are collected. In the noise data, the presence or absence of one or more noise signature of one or more disturbing power supply unit, is detected. Next, lines that are impacted by a noise signature from a common disturbing power supply unit are grouped into one or more groups of common victim lines (G1, G2, G3). Within each group (G1, G2, G3) a disturbing line (d1, d2, d3) holding the common disturbing power supply unit is identified as the line with most severe noise signature from the common disturbing power supply unit.