Diagnostic methods for twisted pair telephone lines based on line data distribution analysis
View Patent ↗Methods and systems for twisted pair telephone line diagnostics based on patterns of line data occurring over time. An observed data distribution is classified as periodic or based on modeled distributions previously determined to correspond to a known line activity, fault type, or fault location. A disruption or parameter value pattern is classified through statistical inference of operational and performance data collected from the line. Where the disruption and/or parameter value(s) correlate with a time the customer is at the customer premises, an inference is made that the line fault causing the disruption is more likely at the CPE than at the Central Office. Where the disruption distribution is classified as being a result of human activities initiated on the line, a fault condition associated with the activity is inferred. Where a disruption pattern is correlated with human initiated plain old telephone service (POTS), a micro-filter problem is inferred for the line.
1. A method of diagnosing a twisted pair telephone line having a DSL transceiver at opposite ends of the line, the method comprising:
identifying a plurality of disruptions in DSL communications on the line through analysis of at least one data parameter sampled over a time period from the line;
generating a temporal distribution of the disruptions over a least a portion of the sampling period;
assessing a correlation between the temporal distribution of the disruptions and one or more reference distributions; and
associating the disruptions with an activity on the line or with a condition of the line corresponding to the one or more reference distributions in response to assessing the correlation.
2. The method of claim 1 , wherein identifying at least one of the disruptions further comprises:
determining an amount of change in at least one parameter collected at two or more consecutive sampling times; and
identifying a disruption in response to the amount of change satisfying a disruption detection threshold.
3. The method of claim 1 , wherein assessing the correlation between the temporal distribution of the disruptions and the one or more reference distributions further comprises determining an observed statistical probability distribution for the disruptions over at least a portion of the sampling time period based on sampling times associated with each disruption.
4. The method of claim 3 , wherein determining the observed statistical probability distribution further comprises determining a probability distribution function (PDF) of intervals of time between consecutive disruptions;
wherein the one or more reference distributions characterize a probability distribution of a human initiated activity on the line; and
wherein associating the disruptions with an activity on the line or with a condition of the line further comprises associating the line disruptions with the human initiated activity.
5. The method of claim 4 , wherein the disruptions comprise retrain events on the line.
6. The method of claim 1 , wherein assessing the correlation between the disruptions distribution and one or more reference distributions further comprises determining a periodicity of the disruptions based on intervals of time between consecutive disruptions;
wherein the one or more reference distributions characterize a periodicity of a machine initiated activity on the line; and
wherein associating the disruptions with an activity on the line or with a condition of the line further comprises associating the disruptions with the machine initiated activity.
7. The method of claim 4 , wherein the initiated line activity is a use of plain old telephone service (POTS) on the line, and wherein associating the disruptions with an activity on the line or with a condition of the line further comprises associating the disruptions with a missing or defective micro-filter on the line.
8. The method of claim 3 , wherein assessing a correlation between the temporal distribution of the disruptions and the one or more reference distributions further comprises:
calculating a goodness of fit (GOF) between a probability distribution function (PDF) and a reference distribution function; and
determining if the GOF satisfies a minimum threshold.
9. The method of claim 8 , wherein calculating the GOF comprises determining an A 2 metric of the Anderson-Darling test, and wherein determining if the GOF satisfies a minimum threshold comprises determining if the A 2 metric corresponds to a threshold level of confidence.
10. The method of claim 1 , further comprising partitioning the sampling period into a plurality of logical time periods including at least one of: a week day, a week night, a weekend day or a weekend night, wherein assessing the correlation between the temporal distribution of the disruptions and one or more reference distributions further comprises determining a separate statistical probability distribution functions (PDF) for the disruptions within each of the logical time periods, and wherein a correlation between the temporal distribution of the disruptions and the one or more reference distributions is assessed separately for each of the logical time periods based on the PDF associated with each logical time period.
11. The method of claim 1 , further comprising:
assessing a quality of the line based on operational or performance data collected from the line; and
adjusting, based on the assessed quality of the line, at least one of:
a disruption detection threshold or a bound on the number of disruptions permitted during portions of the sample time period;
a threshold defining a minimum sufficiency for assessing the correlation between the disruption distribution and the one or more reference distributions; or
a threshold defining a minimum number of observed disruptions.
12. The method of claim 11 , wherein the disruption detection threshold is adjusted to decrease detection sensitivity for a lower quality line or increase detection sensitivity for a higher quality line, and
wherein assessing the quality of the line further comprises at least one of:
assessing channel and noise qualities of the line based on statistics derived from operational data collected over time from at least one of DSL transceivers; or
assessing a stability of the line based on at least one of:
code violation (CV) count,
number of retrains,
forward error correction (FEC) violations,
errored seconds counts, or
severe error seconds (SES).
13. The method of claim 1 , comprises collecting data which includes operational or performance data collected from at least one of DSL transceivers.
14. The method of claim 1 , further comprising:
collecting operational data by polling at least one of DSL transceivers for the value of at least one parameter at known sampling times over the sampling time period, or
collecting performance data by recording a level of counts for at least one parameter at a predetermined sampling frequency over the sampling time period.
15. A system for diagnosing a twisted pair telephone line, the system comprising:
a memory to store values of at least one data parameter sampled from the line;
an analysis module communicatively coupled to the memory to:
identify a plurality of disruptions in DSL communications on the line through analysis of at least one data parameter sampled over a time period from the line;
generate a temporal distribution of the disruptions over a least a portion of the sampling time period;
assess a correlation between the temporal distribution of the disruptions and a reference distribution; and
associate the disruptions with an activity on the line or with a condition of the line corresponding to the reference distribution in response to assessing the correlation.
16. The system of claim 15 , wherein the analysis module is to:
assess the temporal distribution of the disruptions by determining a statistical probability distribution for the disruptions over at least a portion of the sampling time period based on a sampling time associated with each of the disruptions;
determine a statistical probability distribution by determining an observed probability distribution function (PDF) of intervals of time between consecutive ones of the disruptions; and
associate the disruptions with the human initiated activity in response to determining a sufficient correlation between the observed PDF and a reference distribution function characterizing a probability distribution of the human initiated activity on the line.
17. The system of claim 16 , wherein the analysis module is to assess a correlation between the temporal distribution of the disruptions and the reference distribution by calculating a goodness of fit (GOF) between the PDF and a reference distribution function and determining if the GOF satisfies a minimum threshold.
18. The system of claim 17 , wherein the analysis module is to evaluate the quality of the line by at least one of:
assessing channel and noise qualities of the line based on statistics derived from operational data collected over time from at least one of DSL transceivers; or
assessing a stability of the line based on at least one of:
code violation (CV) count,
number of retrains,
forward error correction (FEC) violations,
errored seconds counts, or
severe error seconds (SES).
19. The system of claim 15 , wherein the analysis module is to:
assess a quality of the line based on operational or performance data collected from the line; and
adjust, based on the assessed quality of the line, at least one of:
a disruption detection threshold or a bound on the number of disruptions permitted during portions of the sample period;
a threshold defining a minimum sufficiency for assessing the correlation between the disruption distribution and the reference distribution; or
a threshold defining a minimum number of observed disruptions.
20. At least one computer readable storage medium comprising instructions thereon, that when executed by a processor cause a computer to perform the method of claim 1 .
21. A system for characterizing a physical configuration of a twisted pair telephone line, the system comprising:
a means to store to values of at least one data parameter sampled from the line;
a means to identify a plurality of disruptions in DSL communications on the line through analysis of at least one data parameter sampled over a time period from the line;
a means to generate a temporal distribution of the disruptions over a least a portion of the sampling period;
a means to assess a correlation between the temporal distribution of the disruptions and a reference distribution; and
a means to associate the disruptions with an activity on the line or with a condition of the line corresponding to the reference distribution in response to assessing the correlation.