Systems and methods for identifying defects in local loops
Automatic testing/analysis of local loops of telecommunications networks includes obtaining bits-per-tone data for a local loop of a telecommunications network and generating a bit value string from the bits-per-tone data. The bit value string is then analyzed to determine whether it includes a bit pattern indicative of an impairment of the local loop. Further approaches for automatically testing local loops of telecommunications networks include obtaining attenuation data for multiple tones carried by the local loop and determining whether the attenuation data falls below thresholds for providing a service using the local loop.
1. A computer-implemented method for analyzing local loops of telecommunications networks, the method comprising:
obtaining trace data for a local loop of a telecommunications network;
decomposing the trace data to generate a subsignal, wherein decomposing the trace data to generate the subsignal includes decomposing the trace data into a plurality of initial subsignals via a singular spectral analysis applied to the trace data and combining the plurality of initial subsignals to generate the subsignal, wherein the plurality of initial subsignals includes at least ten initial subsignals;
extracting features from the trace data and the subsignal, wherein the extracted features represent one or more types of information identified from the generation of the subsignal that are useful in identifying defects of the local loop;
providing the features to a classifier, wherein the classifier identifies a class of the local loop based on the features, and wherein the class corresponds to a local loop status indicating a presence of a defect of the local loop, and
transmitting an indicator corresponding to the status, wherein, when received by a computing device, the indicator causes the computing device to display the status and, when a defect is present, to provide access to technical information for resolving the defect.
2. The computer-implemented method of claim 1 , wherein the trace data includes at least one of bits-per-tone data, signal-to-noise ratio data, quiet-line-noise data, or h log data.
3. The computer-implemented method of claim 1 , wherein the subsignal is an empirical orthogonal function.
4. The computer-implemented method of claim 1 , wherein the features include at least one of:
a spectral entropy;
a local binary pattern;
a degree of linearity;
a linear trend;
a degree of non-linearity;
a curvature;
a vanishing point;
a stability of signal mean;
a stability of signal variance;
a median cross;
a fractional differencing order;
a volatility;
a standard deviation of a first derivative;
a peak;
a trough;
a speed provision ratio; and
a line capacity ratio.
5. The computer-implemented method of claim 1 , wherein the classifier includes one or more machine learning models, and wherein providing the features to a classifier includes providing the features to each of the one or more machine learning models.
6. A computing device comprising:
one or more data processors; and
a non-transitory computer-readable storage medium containing instructions which, when executed by the one or more data processors, cause the one or more data processors to perform operations including:
obtaining trace data for a local loop of a telecommunications network;
decomposing the trace data to generate a subsignal, wherein decomposing the trace data to generate the subsignal includes decomposing the trace data into a plurality of initial subsignals via a singular spectral analysis applied to the trace data and combining the plurality of initial subsignals to generate the subsignal, wherein the plurality of initial subsignals includes at least ten initial subsignals;
extracting features from the trace data and the subsignal, wherein the extracted features represent one or more types of information identified from the generation of the subsignal that are useful in identifying defects of the local loop;
providing the features to a classifier, wherein the classifier identifies a class of the local loop based on the features, and wherein the class corresponds to a local loop status indicating a presence of a defect of the local loop; and
transmitting an indicator corresponding to the status, wherein, when received by a computing device, the indicator causes the computing device to display the status and, when a defect is present, to provide access to technical information for resolving the defect.
7. The computing device of claim 6 , wherein the subsignal is an empirical orthogonal function.
8. The computing device of claim 7 wherein the feature includes at least one of:
a spectral entropy, a local binary pattern, a degree of linearity, a linear trend, a degree of nonlinearity, a curvature, a vanishing point, a stability of signal mean, a stability of signal variance, a median cross, a fractional differencing order, a volatility, a standard deviation of a first derivative, a peak, a trough, a speed provision ratio, and a line capacity ratio.
9. The computing device of claim 6 wherein the trace data includes at least one of bits-per-tone data, signal-to-noise ratio data, quiet-line-noise data, or h log data.
10. The computing device of claim 6 wherein the classifier includes one or more machine learning models, and wherein providing the feature to a classifier includes providing the feature to each of the one or more machine learning models.
11. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause a computing device to perform operations including:
obtaining trace data for a local loop of a telecommunications network;
decomposing the trace data to generate a subsignal, wherein decomposing the trace data to generate the subsignal includes decomposing the trace data into a plurality of initial subsignals via a singular spectral analysis applied to the trace data and combining the plurality of initial subsignals to generate the subsignal, wherein the plurality of initial subsignals includes at least ten initial subsignals;
extracting features from the trace data and the subsignal, wherein the extracted features represent one or more types of information identified from the generation of the subsignal that are useful in identifying defects of the local loop;
providing the features to a classifier, wherein the classifier identifies a class of the local loop based on the features, and wherein the class corresponds to a local loop status indicating a presence of a defect of the local loop, and
transmitting an indicator corresponding to the status, wherein, when received by a computing device, the indicator causes the computing device to display the status and, when a defect is present, to provide access to technical information for resolving the defect.
12. The computer-program product of claim 11 , wherein the trace data includes at least one of bits-per-tone data, signal-to-noise ratio data, quiet-line-noise data, or h log data.
13. The computer-program product of claim 11 , wherein the subsignal is an empirical orthogonal function.
14. The computer-program product of claim 11 , wherein the classifier includes one or more machine learning models, and wherein providing the features to a classifier includes providing the features to each of the one or more machine learning models.