Double-ended line probing (DELP) for DSL systems
View Patent ↗Double-ended line probing (DELP) techniques are described herein that enable a loop configuration to be identified based on mathematically simple error functions, which are both time-efficient and less sensitive to measurement errors. A DELP algorithm uses readily available data about the channel that is generated prior to the modem operation (e.g., theoretical loop models) and other available data (e.g., that provided by loop diagnostic modules) to estimate the loop configuration including the line length and gauge, as well as the number of bridge taps and their gauges, lengths, and locations.
1. A method for double-ended line probing (DELP) on a copper telephone line having up to two bridge taps located thereon, comprising:
measuring an overall channel insertion loss magnitude on the copper telephone line;
generating a bridge tap error function #1 which includes the measured channel insertion loss magnitude measured in the measuring act and estimated insertion loss magnitudes of the bridge taps for all possible bridge tap gauges and lengths;
estimating a gauge and length for each bridge tap that minimizes the bridge tap error function #1 generated in the generating act;
deriving the insertion loss for the line with no bridge taps by dividing the overall channel insertion loss measured in the measuring act by a product of insertion loss magnitudes associated with the estimated bridge tap length and gauge estimated in the estimating act; and
estimating a first approximation of a length and gauge for the line based on the insertion loss of the line with no bridge taps derived in the deriving act;
thereby transforming the measured overall channel insertion, loss magnitude into physical and topological parameters of the copper telephone line, including both the estimated length and gauge of the telephone line together with estimated bridge tap length and gauge thereof.
2. The method of claim 1 , further comprising:
generating a line error function for all possible line gauges and lengths, based on the estimated locations, gauges and lengths of the bridge taps, and the approximation of the length and gauge for the line; and
calculating a second approximation of line length and gauge by minimizing the line error function over all line gauges and lengths.
3. The method of claim 2 , wherein the line error function is
error=∫| H Data ( f )− H Test ( f )| df
where H Data (f) is a measured overall channel insertion loss magnitude, and H Test is a theoretical overall channel insertion loss magnitude of a line with bridge taps having their length, gauge and location equal to the estimated bridge taps' length, gauge and location, and once the line length and gauge that minimizes the line error function is identified, the second approximation of line length and gauge is also identified.
4. The method of claim 2 , wherein in response to determining that a difference between a measured overall insertion loss magnitude and an estimated overall insertion loss magnitude is greater than a pre-defined error threshold, the method includes iterating a number of times to improve parameter estimation results.
5. The method of claim 1 , further comprising:
generating a bridge tap error function #2 for all possible bridge taps locations, based on the estimated gauges and lengths of the bridge taps and the approximation of the length and gauge for the line; and
estimating the location of each bridge taps by minimizing the bridge taps error function #2 over all bridge tap locations.
6. The method of claim 1 , further comprising:
deriving a channel insertion loss magnitude for the line without the bridge taps, by dividing a measured overall channel insertion loss magnitude by a product of channel insertion loss magnitudes associated with the estimated gauges and lengths of the bridge taps; wherein the estimating a first approximation of a length and gauge for the line is based on the derived channel insertion loss magnitude and a theoretical loop model that associates channel insertion loss magnitude for the line and line length and gauge.
7. The method of claim 1 , wherein the bridge tap error function #1 is
error
=
∫
Bin_start
Bin_end
ⅆ
2
ⅆ
f
2
(
H
Data
(
f
)
H
Tap
1
(
f
)
⨯
H
Tap
2
(
f
)
)
ⅆ
f
,
where Bin_start is a start of channel frequency range, and Bin_end is an end of channel frequency range, H Data (f) is a measured overall channel insertion loss magnitude, and H Tap 1 (f) and H Tap 2 (f) are channel insertion loss magnitudes of the bridge taps from a theoretical loop model and, once the bridge tap channel insertion loss magnitudes that minimize the error function #1 are identified within the model, the estimated length and gauge of those bridge taps are also identified within the model.
8. The method of claim 5 , wherein the bridge tap error function #2 is
error=∫| H One — Tap ( f )− H Test ( f )| df
where H One — Tap is an approximation of channel insertion loss magnitude of a line with only one bridge tap, and H Test is a theoretical overall channel insertion loss magnitude of a single tap line with line length and gauge equal to the first approximation of length and gauge for the line, and bridge tap length and gauge equal to those associated with H One — Tap , and once the bridge tap location that minimizes the error function #2 is identified, the estimated bridge tap location is also identified.
9. The method of claim 8 , wherein
H One — Tap =H Data ( f )/ H Tapj ( f )
where H Data (f) is a measured overall channel insertion loss magnitude, and H Tapj (f) is a channel insertion loss magnitude for one of the bridge taps estimated using error equation #1.
10. A means for double-ended line probing (DELP) on a copper telephone line having up to two bridge taps located thereon, comprising:
means for measuring an overall channel insertion loss magnitude on the copper telephone line;
means for generating a bridge tap error function #1 which includes the measured channel insertion loss magnitude measured by the means for measuring and estimated insertion loss magnitudes of the bridge taps for all possible bridge tap gauges and lengths;
means for estimating a gauge and length for each bridge tap that minimizes the bridge tap error function #1 generated by the means for generating;
means for deriving the insertion loss for the line with no bridge taps by dividing the overall channel insertion loss measured by the means for measuring by a product of insertion loss magnitudes associated with the estimated bridge tap length and gauge estimated by the means for estimating; and
means for estimating a first approximation of a length and gauge for the line based on the insertion loss of the line with no bridge taps derived by the means for deriving; thereby transforming the measured overall channel insertion loss magnitude into physical and topological parameters of the copper telephone line, including both the estimated length and gauge of the telephone line together with estimated bridge tap length and gauge thereof.
11. The means of claim 10 , further comprising:
means for deriving a channel insertion loss magnitude for the line without the bridge taps, by dividing a measured overall channel insertion loss magnitude by a product of channel insertion loss magnitudes associated with the estimated gauges and lengths of the bridge taps; wherein the estimating a first approximation of a length and gauge for the line is based on the derived channel insertion loss magnitude and a theoretical loop model that associates channel insertion loss magnitude for the line and line length and gauge.
12. The method of claim 10 , wherein the bridge tap error function #1.is
error
=
∫
Bin_start
Bin_end
ⅆ
2
ⅆ
f
2
(
H
Data
(
f
)
H
Tap
1
(
f
)
⨯
H
Tap
2
(
f
)
)
ⅆ
f
,
where Bin_start is a start of channel frequency range, and Bin_end is an end of channel frequency range, H Data (f) is a measured overall channel insertion loss magnitude, and H Tap 1 (f) and H Tap 2 (f) are channel insertion loss magnitudes of the bridge taps from a theoretical loop model and, once the bridge tap channel insertion loss magnitudes that minimize the error function #1 are identified within the model, the estimated length and gauge of those bridge taps are also identified within the model.
13. The means of claim 10 , further comprising:
generating a bridge tap error function #2 for all possible bridge tap locations, based on the estimated gauges and lengths of the bridge taps and the approximation of the length and gauge for the line; and
estimating the location of each bridge tap by minimizing the bridge tap error function #2 over all bridge tap locations.
14. The means of claim 13 , wherein the bridge tap error function #2 is
error=∫| H One — Tap ( f )− H Test ( f )| df
where H One — tap is an approximation of channel insertion loss magnitude of a line with only one bridge tap, and H Test is a theoretical overall channel insertion loss magnitude of a single tap line with line length and gauge equal to the first approximation of length and gauge for the line, and bridge tap length and gauge equal to those associated with H One — tap , and once the bridge tap location that minimizes the error function #2 is identified, the estimated bridge tap location is also identified.
15. The means of claim 14 , wherein
H One — Tap =H Data ( f )/ H Tapj ( f )
where H Data (f) is a measured overall channel insertion loss magnitude, and H Tapj (f) is a channel insertion loss magnitude for one of the bridge taps estimated using error equation #1.
16. The method of claim 13 , further comprising:
generating a line error function for all possible line gauges and lengths, based on the estimated locations, gauges and lengths of the bridge taps, and the approximation of the length and gauge for the line; and
calculating a second approximation of line length and gauge by minimizing the line error function over all line gauges and lengths.
17. The means of claim 16 , wherein the line error function is
error=∫| H Data ( f )− H Test ( f )| df
where H Data (f) is a measured overall channel insertion loss magnitude, and H Test is a theoretical overall channel insertion loss magnitude of a line with bridge taps having their length, gauge and location equal to the estimated bridge taps' length, gauge and location, and once the line length and gauge that minimizes the line error function is identified, the second approximation of line length and gauge is also identified.
18. The means of claim 16 , wherein in response to determining that a difference between a measured overall insertion loss magnitude and an estimated overall insertion loss magnitude is greater than a pre-defined error threshold, the method includes iterating a number of times to improve parameter estimation results.