U.S. patent application number 12/276682 was filed with the patent office on 2009-10-22 for repeaterless backhaul.
Invention is credited to William L. Betts, Gordon F. Bremer, Philip J. Kyees.
Application Number | 20090262911 12/276682 |
Document ID | / |
Family ID | 29408281 |
Filed Date | 2009-10-22 |
United States Patent
Application |
20090262911 |
Kind Code |
A1 |
Bremer; Gordon F. ; et
al. |
October 22, 2009 |
Repeaterless Backhaul
Abstract
Using the concepts of inverse multiplexing, repeaterless
backhaul can be provided on loaded or unloaded loops. Inverse
multiplexing can effectively support a large bit rate capacity
between the conversion equipment and the CO-side equipment by
utilizing a large enough number of potentially low capacity loops.
As the distance that the data has to be backhauled increases, the
bit rate capacities of the loops generally decrease. However, this
decrease in the bit rate capacities of the loops can be compensated
for by using inverse multiplexing to gather together enough loops
to meet the data rate requirements for backhauling customer data
given various service level and contention criteria.
Inventors: |
Bremer; Gordon F.; (Largo,
FL) ; Betts; William L.; (St. Petersburg, FL)
; Kyees; Philip J.; (Largo, FL) |
Correspondence
Address: |
WOODCOCK WASHBURN LLP
CIRA CENTRE, 12TH FLOOR, 2929 ARCH STREET
PHILADELPHIA
PA
19104-2891
US
|
Family ID: |
29408281 |
Appl. No.: |
12/276682 |
Filed: |
November 24, 2008 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10434864 |
May 8, 2003 |
7471777 |
|
|
12276682 |
|
|
|
|
60379124 |
May 8, 2002 |
|
|
|
60379026 |
May 8, 2002 |
|
|
|
60379040 |
May 8, 2002 |
|
|
|
60379030 |
May 8, 2002 |
|
|
|
60379029 |
May 8, 2002 |
|
|
|
60379038 |
May 8, 2002 |
|
|
|
60379041 |
May 8, 2002 |
|
|
|
60379028 |
May 8, 2002 |
|
|
|
60379142 |
May 8, 2002 |
|
|
|
Current U.S.
Class: |
379/93.05 |
Current CPC
Class: |
H04L 5/16 20130101; H04L
12/66 20130101 |
Class at
Publication: |
379/93.05 |
International
Class: |
H04M 11/00 20060101
H04M011/00 |
Claims
1. A method of providing repeaterless backhaul, the method
comprising the steps of: inverse multiplexing data over a plurality
of loops; and selecting enough of the plurality of loops to meet a
bandwidth requirement over a distance the data is to be
backhauled.
2. The method of claim 1, wherein at least one of the plurality of
loops are repeaterless loops.
3. The method of claim 1, further comprising: receiving the data
over a first loop according to a first modulation method; and
transmitting the data over the plurality of loops according to a
second modulation method.
4. The method of claim 1, wherein inverse multiplexing comprises
inverse multiplexing data across a first pair of wires, a second
pair of wires, and a third pair of wires, such that data is
communicated in a forward direction on the first pair of wires, in
a reverse direction on the second pair of wires, and in a direction
selectable from the forward direction and the reverse direction on
the third pair of wires.
5. The method of claim 4, wherein the direction selectable from the
forward direction and the reverse direction changes between the
forward direction and the reverse direction based on data
demand.
6. The method of claim 1, wherein selecting comprises selecting
loops in accordance with a contention ratio.
7. A method of providing repeaterless backhaul, the method
comprising: selecting a plurality of loops, wherein a count of
loops in the plurality of loops is based on a backhaul bandwidth
requirement and a backhaul distance requirement; and inverse
multiplexing data over the plurality of loops.
8. The method of claim 7, wherein at least one of the plurality of
loops are repeaterless loops.
9. The method of claim 7, further comprising: receiving the data
over a first loop according to a first modulation method; and
transmitting the data over the plurality of loops according to a
second modulation method.
10. The method of claim 9, wherein the first modulation method is a
digital subscriber line (DSL) modulation method.
11. The method of claim 9, wherein the second modulation method is
a digital subscriber line (DSL) modulation method.
12. The method of claim 9, wherein the first loop carries
communication of plain old telephone service (POTS) data and
digital subscriber line (DSL) data.
13. The method of claim 7, wherein inverse multiplexing comprises
inverse multiplexing data across a first pair of wires, a second
pair of wires, and a third pair of wires, such that data is
communicated in a forward direction on the first pair of wires, in
a reverse direction on the second pair of wires, and in a direction
selectable from the forward direction and the reverse direction on
the third pair of wires.
14. The method of claim 13, wherein the direction selectable from
the forward direction and the reverse direction changes between the
forward direction and the reverse direction based on data
demand.
15. The method of claim 7, wherein selecting enough comprises
selecting loops in accordance with a contention ratio.
16. A device for providing repeaterless backhaul, the device
comprising: an first interface to a plurality of loops; and an
inverse multiplexer in communication with the interface, wherein
the inverse multiplexer inverse multiplexes data over the plurality
of loops, wherein a count of loops in the plurality of loops is
based on a backhaul bandwidth requirement and a backhaul distance
requirement.
17. The device of claim 16, wherein the plurality of loops are
repeaterless loops.
18. The device of claim 16, further comprising: a second interface
in communication with the inverse multiplexer, wherein the second
interface receives the data over a first loop according to a first
modulation method, and wherein the first interface transmits the
data over the plurality of loops according to a second modulation
method.
19. The device of claim 16, where the inverse multiplexer inverse
multiplexes data across a first pair of wires, a second pair of
wires, and a third pair of wires, such that data is communicated in
a forward direction on the first pair of wires, in a reverse
direction on the second pair of wires, and in a direction
selectable from the forward direction and the reverse direction on
the third pair of wires.
20. The device of claim 16, further comprising a multiplexer, in
communication with the first interface, to multiplex data from the
plurality of loops.
Description
CROSS-REFERENCE TO RELATED PATENT APPLICATIONS
[0001] This application is a divisional of U.S. patent application
Ser. No. 10/434,864 filed on May 8, 2003, incorporated by reference
in its entirety herein, which claims the benefit of the following
U.S. provisional applications, each filed on May 8, 2002, each
incorporated by reference in its entirety herein:
[0002] 60/379,124--"Loaded Loop DSL Service and Business
Method"
[0003] 60/379,026--"Loaded Loop DSL Modulation"
[0004] 60/379,040--"Automatic Selection of Loaded Loop DSL or
Standard DSL Service"
[0005] 60/379,030--"Loaded Loop DSL Loop Intercession System"
[0006] 60/379,029--"Automatic Switching Between Loaded Loop DSL and
POTS"
[0007] 60/379,038--"Multiple POTS Services on a Loaded Loop"
[0008] 60/379,041 "Utilization of More Than One Loaded Loop for DSL
Service"
[0009] 60/379,028--"Data Communication Over Loaded Loops
Simultaneous with POTS Call"
[0010] 60/379,142--"Indirect Loaded Loop DSL."
[0011] Furthermore, this application is related to U.S. Pat. Nos.
7,289,610 and 7,272,215, each incorporated by reference in its
entirety herein.
TECHNICAL FIELD
[0012] The present invention generally is related to
telecommunications and, more particularly, is related to a system
and method for improving the delivery of digital subscriber line
(DSL) service.
BACKGROUND
[0013] Digital Subscriber Line or Loop (DSL) communication
technologies have been adopted by telephone service providers as a
way of extending digital service to customer premises (CP) such as
homes and offices. The advent of digital communication technology
has resulted in an evolutionary change to communication systems as
the facilities of switches and trunks in the networks of
telecommunications service providers were converted first from
analog to digital. Next, consumers wanted digital access to these
digital capabilities in the network facilities of service
providers. However, delivering digital services over the local loop
or subscriber line facilities to cover what is often colloquially
called "the last mile" to the customer premises has been more of a
challenge to provision. While various mechanisms have been used to
deliver digital services to customer premises, making major changes
to the wiring plant that feeds subscribers generally is still
prohibitively costly. For customers located near a central office
(CO) or close to a digital loop carrier (DLC) system, with the
generally corresponding short cable wiring runs, DSL service is
often available.
[0014] However, DSL capabilities still are not available to many
customers located at farther reaches from central office switches
and/or digital multiplexers such as a DLC. Furthermore, the
historical telephone wiring plant feeding many customer locations
was designed and optimized for the analog voice frequency
communications of plain old telephone service (POTS) primarily
found in the 0 to 4 KHz range. (One skilled in the art will be
aware that the common bandwidth for unloaded POTS loops is
primarily found in the 0 to around 4 KHz range, while the common
bandwidth for loaded POTS loops is primarily found in the 0 to
around 3.4 KHz range. One skilled in the art will be aware of these
actual bandwidth differences of loaded and unloaded loops in
carrying native POTS communication even though the POTS baseband is
commonly referred to as a 0-4 KHz POTS baseband. One of ordinary
skill in the art will be aware that such a reference is not
completely accurate for loaded loops, but is a useful shorthand
when discussing the POTS baseband configurations.)
[0015] Historically, telephone companies often found it
advantageous to install inductors or load coils on many local loops
to optimize performance of the loops in carrying POTS voice
communication. Generally, the load coils or inductors were
installed in series at various points along the telephone local
loop. On a properly designed local loop, load coils generally are
placed on subscriber loops that are greater than or equal to 18
Kft. in length. The load coils commonly used by the Regional Bell
Operating Companies (RBOCs) have 88 milli-Henrys as the standard
nominal inductance value for the coils. In general, load coils are
spaced along a subscriber loop beginning at approximately 3 Kft.
from a line card in a CO switch or DLC chassis with additional
coils generally spaced along the loop approximately each 6 Kft.
thereafter. The customer end portion of a local loop generally is
allowed to have lengths ranging from 3 Kft. to 12 Kft. beyond the
last load coil. In general, the local loop design rules used by the
RBOCs specify that three or more load coils should be used on loops
that are 18 Kft. or longer in length. In some special assembly
situations, such as but not limited to analog POTS loops used as
trunks for a customer's PBX, the RBOCs may use load coils on loops
as short as 15 Kft. in length with a minimum of two load coils.
[0016] Essentially, adding an inductor in series results in the
creation of a low pass filter. While the low-pass filtering of
these load coils improves performance in the 0 to around 3 KHz base
bandwidth of an analog POTS interface, the filtering results in
detrimental effects (primarily attenuation) on the higher frequency
signals above 3 KHz that generally are used in DSL technologies.
Unfortunately, the problem is not solved simply by getting the
service provider to remove the load coils on each loop. While such
an action certainly solves the technical limitations of load coils
on DSL performance, economically it is an expensive process to
remove the load coils. Furthermore, removal of the loading coils re
introduces the voice-band degradations that the coils were
introduced to overcome. As a result, the service provider often
cannot justify the costs of basically custom re-engineering each of
the multitude of subscriber lines to remove load coils in order to
earn the additional revenues from offering DSL service. Removing
load coils generally would involve identifying the location of all
of the load coils on a subscriber loop and sending a technician to
each location to take the load coil out of the subscriber line
circuit. Just sending the technician to each location would be
costly enough. However, the physical process of removing load coils
can create additional problems. For instance, most cables in the
underground are pulp insulated such that wire pairs can be easily
damaged as a result of a technician or cable splicer working on the
splice to locate the wire pair affected by load coils. Obviously,
damage to other pairs may knock out phone service to existing
customers.
[0017] In addition, often the databases and records of service
providers are incomplete and/or inaccurate in keeping up with the
location of all the load coils that were installed on a particular
subscriber loop over the years. Thus, in some cases various
transmission line tests (such as, but not limited to tests
performed by a time-domain reflectometer or TDR) might have to be
performed to determine the distance along a subscriber loop
transmission line at which there are changes in the characteristic
impedance of the transmission line indicating potential items such
as, but not limited to, load coils, junction splices, bridge taps,
and/or connection points.
[0018] Because an impedance mismatch in a transmission line causes
at least part of the energy from propagating electromagnetic
signals to be reflected or echoed back in the opposite direction of
the original propagation, a TDR and other types of test equipment
generally can be used to send signals down a transmission line and
measure the amount of time before a signal reflection or echo is
received at the test equipment. This time measurement together with
the estimated speed of propagation of the electromagnetic wave in
the transmission line medium can be used to provide an estimate of
the distance along the transmission line (such as a subscriber
loop) where impedance mismatches occur. In general, telephone
companies (or telcos) maintain computerized or paper plat records
showing the location of telco facilities such as, but not limited
to, wires, splice points, cross-connects, and DLCs used in
delivering service to residential and commercial areas. The
transmission line distances provided by a TDR or other test
equipment for the potential location of impedance mismatches, which
might be caused by load coils, would have to be used to estimate
the approximate geographic location of a load coil based on the
potentially inaccurate service provider records showing the wiring
path for the transmission line from the central office or DLC to
the customer premises. Obviously, such activities of identifying
load coils and possibly having a technician physically track down
the path followed by a subscriber loop transmission line can be
costly. As a result of these load coil issues, either some
customers are not offered DSL service at all or the price of the
service is higher than it should be because of the increased costs
of removing load coils. Thus, service providers are not able to
offer DSL service to a relatively larger number of potential
subscribers because of the load coil issue. Improving this load
coil problem would increase the number of customers and associated
revenues available to the service provider.
[0019] In addition, subscriber loops normally run through various
other facilities in connecting a customer premises to a line card
in a central office switch or in a digital multiplexer such as a
DLC. Often telephone wiring is run in groups of large multi pair
cables from a connection co-located with the line cards to a splice
point, junction terminal, or cross-connect point. The cross-connect
point generally is an unpowered box where technicians can
cross-connect the wires leading to a customer premises with the
appropriate wires leading back to the line cards in a switch or
DLC. Often the portion of a local loop transmission line from a
cross-connect box back to a line card is known as the F1 or feeder
portion of a local loop, while the portion of a local loop
transmission line from the cross-connect to the customer premises
is known as the F2 or distribution portion of a local loop.
Normally, the cross-connect box uses various mechanical
technologies (such as but not limited to various punch-down block
technologies) that are common in telephone wiring to simplify a
technician's work in connecting the two portions of a subscriber
loop. Unlike a digital loop carrier (DLC) cabinet, which generally
is provided with power from the central office (and/or other
sources) to enable the operation of the electronic devices of the
line cards and multiplexing equipment, cross-connect boxes and/or
cabinets generally are not provided with power other than the
powering delivered over the POTS interface of each in-service POTS
loop that provides for basic POTS functionality powering to a
customer premises. This power on a POTS loop is designed for
powering POTS analog phones with basic functionality (such as, but
not limited to, dial tone) at the customer premises and generally
does not provide a significant amount of excess power that could be
siphoned off to power other types of electronic digital
communications equipment. Often analog phones with POTS interfaces
that offer more functionality such as a speaker phone or memory
need additional power from an AC outlet or battery at the customer
premises because the POTS interface does not provide enough power
to meet the needs of these additional electronic functions.
[0020] In providing DSL service, often the network-side or CO-side
of the DSL line is terminated in a DSLAM (Digital Subscriber Line
Access Multiplexer) that usually is capable of supporting multiple
DSL loops. One skilled in the art will be aware that a DSLAM
normally comprises a plurality of DSL modems and some statistical
multiplexing concentration equipment. However, such DSLAM equipment
normally needs a reasonable amount of power and is usually placed
in locations where power is readily available such as a central
office (CO) or DLC cabinet. As cross-connect boxes generally do not
have power available for powering active electronics, DSLAMs are
not placed in cross-connect boxes. Furthermore, cross-connect boxes
generally are not large enough to encompass significant amounts of
additional electronic equipment in contrast to the relatively
larger cabinets containing DLCs. Thus, normal deployment of DSLAMs
for providing DSL service to customers does not place DSLAMs in
cross-connect boxes at least because cross-connect boxes generally
do not have a ready source of sufficient power and cross-connect
boxes generally are not large enough for holding the DSLAM
equipment.
[0021] Given these and other limitations of the wiring cable plant
that was often originally installed many years ago to just provide
basic POTS, new innovations that increase the availability and
lower the total costs of delivering digital subscriber line (DSL)
service provide benefits that can allow more consumers to obtain a
reasonable digital service access line at an affordable price
point.
SUMMARY
[0022] Using the concepts of inverse multiplexing, repeaterless
backhaul can be provided on loaded or unloaded loops. Generally,
extending the transmission line distance between two communication
devices lowers the potential channel capacity of the transmission
line, other things being equal. Repeaters are one solution to this
problem by keeping the data rate high by basically lowering the
distance over which digital signals have to propagate before a
clean copy of the information can be regenerated at a repeater. In
general, backhauling requires among other things meeting some
minimum data rate requirements for serving the acceptable
contention ratios of the backhaul link. The preferred embodiments
of the present invention include inverse multiplexing that can
effectively support a large bit rate capacity between the
conversion equipment and the CO-side equipment by utilizing a large
enough number of potentially low capacity loops. As the distance
that the data has to be backhauled increases, the bit rate
capacities of the loops generally decrease. However, this decrease
in the bit rate capacities of the loops can be compensated for by
using inverse multiplexing to gather together enough loops to meet
the data rate requirements for backhauling customer data given
various service level and contention criteria. Thus, the inverse
multiplexing of the preferred embodiments of the present invention
also helps to resolve the repeater problem for backhauling the DSL
data (and potentially the digitized POTS).
[0023] Other systems, methods, features, and advantages of the
present invention will be or become apparent to one with skill in
the art upon examination of the following drawings and detailed
description. It is intended that all such additional systems,
methods, features, and advantages be included within this
description, be within the scope of the present invention, and be
protected by the accompanying claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The invention can be better understood with reference to the
following drawings. The components in the drawings are not
necessarily to scale, emphasis instead being placed upon clearly
illustrating the principles of the present invention. Moreover, in
the drawings, like reference numerals designate corresponding parts
throughout the several views.
[0025] FIG. 1 is a block diagram of a single subscriber indirect
DSL configuration over a loaded subscriber loop using Paradyne's
ReachDSL technology.
[0026] FIG. 2 is a block diagram of a single subscriber indirect
DSL configuration over a loaded subscriber loop using standard ADSL
technology that cannot work with any load coils.
[0027] FIG. 3 is a block diagram of a single subscriber indirect
DSL configuration over a loaded subscriber loop without supporting
a native POTS interface on the subscriber loop.
[0028] FIG. 4 is a block diagram of a multi-subscriber indirect DSL
configuration over at least one loaded subscriber loop using
Paradyne's ReachDSL technology.
[0029] FIG. 5 is a block diagram of a multi-subscriber indirect DSL
configuration over at least one loaded subscriber loop using
standard ADSL technology that cannot work with any load coils.
[0030] FIG. 6 is a block diagram of a multi-subscriber indirect DSL
configuration over at least one loaded subscriber loop without
supporting native POTS interfaces to the customer premises.
[0031] FIG. 7 is a more detailed block diagram showing an
embodiment of conversion equipment for a single subscriber indirect
DSL configuration over a loaded subscriber loop using Paradyne's
ReachDSL technology.
[0032] FIG. 8 is a more detailed block diagram showing an
embodiment of conversion equipment for a multi-subscriber indirect
DSL configuration over a loaded subscriber loop using Paradyne's
ReachDSL technology.
[0033] FIG. 9 is a block diagram of loaded F1 feeder loops as well
as loaded and unloaded F2 distribution loops connected to a
cross-connect cabinet.
[0034] FIG. 10 is a block diagram of unloaded F1 feeder loops and
unloaded F2 distribution loops connected to a cross-connect
cabinet.
[0035] FIG. 11 is a block diagram showing the co-location of a
cross-connect cabinet and conversion equipment that uses loaded
loops in providing different types of DSL and POTS service to
multiple customers.
[0036] FIG. 12 is a block diagram showing more detail of conversion
equipment in supporting multiple customers with DSL service.
[0037] FIG. 13 is a block diagram shown one approach that attempts
to successfully provide DSL service from a cross-connect box over
loaded loops.
[0038] FIG. 14 is a block diagram showing one approach that
attempts to successfully provide DSL service from a cross-connect
box over unloaded loops.
[0039] FIG. 15 is a block diagram showing an indirect DSL approach
that successfully provides DSL service from a cross-connect box
over loaded loops and/or unloaded loops for backhaul.
[0040] FIG. 16 is a block diagram showing the potential wiring
problems of the approaches in FIGS. 13 and 14 that create
reliability problems for lifeline POTS service.
[0041] FIG. 17 is a block diagram showing the potential connections
for bridging the DSL service onto subscriber loops without creating
the POTS reliability problems.
DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
[0042] Various types of modulation and/or duplexing techniques can
be used to overcome the limitations of load coils. By their nature,
the low pass, frequency-dependent filtering of load coils creates
serious adverse effects on duplexing strategies that depend at
least in part on frequency as well as on multiplexing strategies
that depend on frequency separation to utilize a subscriber loop
both for a basic native POTS interface in the 0-4 KHz POTS baseband
together with the associated POTS signaling (such as but not
limited to, POTS ringing, call waiting, calling line ID, and/or
dual-tone multi-frequency (DTMF) or pulse/rotary dialing) and for
DSL service. One skilled in the art will be aware of the
differences between a native POTS interface on a subscriber access
line or loop and a derived or synthesized voice call functionality
that may be provided through a local POTS interface off of customer
premises equipment (CPE) that connects to one or more digital
channels or digital media on a digital subscriber access line or
loop. One non-limiting example of such a derived or synthesized
voice call functionality is the circuit-switched speech capability
available from an analog POTS port on an ISDN BRI terminal adapter
(TA) that utilizes a 64 kbps ISDN B-channel to carry digital pulse
code modulation (PCM) samples over the digital subscriber access
loop or line back to the line card. Other more current synthesized
or derived digital voice technologies often utilize compressed
and/or packetized encoding of human voice instead of the 64 kbps or
56 kbps DSOs. Many but not all customers prefer DSL service to be
offered on the same access line/loop (or lines/loops) that supports
a basic native POTS interface in the 0-4 KHz baseband because many
customers often view basic POTS service as a lifeline of the bare
minimum quality of connectivity that is expected to be available in
all but the most catastrophic emergency conditions such as an
earthquake or hurricane. Thus, a solution that provides faster
digital service over a loaded subscriber loop together with a
capability to support an analog POTS interface is important.
[0043] In addition, with reference to all FIGS. 1-17, one skilled
in the art will be aware that equipment labeled as central office
(CO) equipment generally is CO-side, service provider-side, or
network-side equipment, which may be located in other reasonable
network-side concentration locations (such as but not limited to a
DLC or a cross-connect box as will be described further) instead of
just being limited to deployment in a CO. In addition, one skilled
in the art will be aware that the terminology of CO-side, service
provider-side, and network-side is commonly used to differentiate
one portion of an interface's functionality from another portion of
an interface's functionality that may be called in different
contexts by terms such as, but not limited to, customer premises
(CP)-side, user-side, or subscriber-side.
[0044] The patent application with attorney docket number
61607-1780, entitled "Digital Subscriber Line Service Over Loaded
Loops", and filed the same day is incorporated by reference in its
entirety herein and describes some techniques for providing DSL
service over loaded loops. The preferred embodiments of the present
invention described herein extend the ability to offer DSL service
over loaded loops and unloaded loops. One non-limiting solution to
the problem of loaded loops is to add some conversion equipment to
a subscriber loop to allow DSL to operate over the loaded loop.
[0045] FIG. 1 shows a potential configuration of adding conversion
equipment to a loaded subscriber loop. In FIG. 1, a ReachDSL modem
1011 and one or more analog POTS phone(s) 1013 or other type of
POTS CPE are connected to a portion of a subscriber loop with one
load coil 1051 (or no load coils). The segment of the subscriber
loop between conversion equipment 1550 and central office equipment
1560 (such as a line card in a switch or DLC) has at least one load
coil arbitrarily shown as load coils 1651, 1653, and 1655. In
effect, conversion equipment 1550 has been inserted in the
subscriber loop between load coil 1051 and load coil 1651 to create
two different segments of the subscriber loop. DSL service over the
subscriber loop segment between conversion equipment 1550 and
central office equipment 1560 could be provided in a non-limiting
case using the techniques described in the patent application with
attorney docket number 61607-1780, entitled "Digital Subscriber
Line Service Over Loaded Loops."
[0046] Furthermore, Paradyne's ReachDSL modem technology will work
over unloaded subscriber loops (or unloaded segments) that include
a single load coil with the loop segment length being up to
approximately 15 Kft. Thus, using Paradyne's ReachDSL technology
between ReachDSL modem 1011 and conversion equipment 1550 can
provide DSL service over that portion or segment of the local loop.
Therefore, the conversion equipment 1550 effectively segments or
divides a subscriber loop to allow both POTS service to be provided
between PSTN 1950 and analog phone(s) 1013 as well as DSL service
to be provided between data network 1960 and ReachDSL modem 1011.
One skilled in the art will be aware that throughout FIGS. 1-17,
the PSTN generally has historically been a circuit-switched
network, while data networks have tended to be packet-switched
networks utilizing statistical multiplexing. However, one skilled
in the art will be aware of many methods for interconnecting
circuit-switched and packet-switched networks and of various
industry trends have caused a convergence of various historically
separate networking technologies. Thus, although the PSTN is
commonly thought of as a circuit-switched network and the data
network is thought of as a packet network, the preferred
embodiments of the present invention are not to be limited to those
common network architectures.
[0047] Referring to FIG. 2, normal ADSL (Asymmetric Digital
Subscriber Line) modems such as ADSL modem 2011 do not reliably
work with any load coils on a subscriber loop. Thus, using standard
ADSL technology, ADSL modem 2011 and analog POTS phone(s) 2013 are
connected over an unloaded loop 2023 segment or portion to
conversion equipment 2550. Conversion equipment 2550 then is
further connected to central office equipment 2560 over a loaded
loop segment comprising at least one load coil arbitrarily shown as
load coils 2651, 2653, and 2655. In effect, conversion equipment
2550 has been inserted in the subscriber loop between load coil
1051 and load coil 2651 to create two different segments of the
subscriber loop. DSL service over the subscriber loop segment
between conversion equipment 2550 and central office equipment 2560
could be provided in a non-limiting case using the techniques
described in the patent application with attorney docket number
61607-1780, entitled "Digital Subscriber Line Service Over Loaded
Loops." DSL service over the loop segment between conversion
equipment 2550 and ADSL modem 2011 would use normal ADSL
techniques. Using such a configuration, POTS service can be
provided between phone(s) 2013 and PSTN 2950, and digital data
service can be provided between ADSL modem 2011 and data network
2960 so long as the conversion equipment 2550 converts the digital
signal between normal ADSL and other techniques that are capable of
operating over loaded loops (or loop segments).
[0048] FIG. 3 shows the special case where a customer is not also
using the same subscriber line for basic analog POTS. In FIG. 3,
ReachDSL modem 3011 is connected to a segment of a subscriber loop
containing at most one load coil 3051 that further connects to
conversion equipment 3550. Conversion equipment 3550 is connected
to central office equipment 3560 over a segment of a subscriber
loop having at least one load coil arbitrarily shown as load coils
3651, 3653, and 3655. As was the case in FIG. 1, the conversion
equipment 3550 has in effect been inserted in the subscriber loop
between load coil 3051 and load coil 3651 to create two different
segments of the subscriber loop. DSL service over the subscriber
loop segment between conversion equipment 3550 and central office
equipment 3560 could be provided in a non-limiting case using the
techniques described in the patent application with attorney docket
number 61607-1780, entitled "Digital Subscriber Line Service Over
Loaded Loops." Because the customer does not use the subscriber
line to carry a basic POTS interface in the 0-4 KHz baseband, the
ReachDSL modem 3011 communicates with data network 3960, and no
PSTN is shown. One skilled in the art will be aware of using
various digitized and/or packetized voice technologies together
with an appropriate gateway connecting the data network 3960 to the
PSTN to provide a customer with PSTN access through the ReachDSL
modem 3011.
[0049] While the configurations of FIGS. 1-3 certainly will work to
provide DSL service by inserting conversion equipment on each
subscriber line that has a problem with load coils, those solutions
are not necessarily the most efficient solutions when larger
numbers of local loops in a common geographic customer service area
have problems with load coils. Certainly, multiple instantiations
of the single loop configurations of FIGS. 1, 2, and 3 could be
used to support large numbers of DSL subscribers with loaded loops.
However, such a configuration does not take advantage of
concentrations when a relatively large number of DSL subscribers on
loaded loops are located in close proximity to be served out of the
same wiring concentration centers and facilities in the service
provider's network. Thus, instead of just scaling up the single
loaded subscriber loop configurations of FIGS. 1, 2, and 3, various
multiplexing strategies can be used to gain some efficiencies when
supporting a larger number of DSL customers. FIGS. 4, 5, and 6 show
how multiplexing might be used in the case of supporting multiple
customers (arbitrarily three customers in FIGS. 4, 5, and 6) for
the loaded loop scenarios that correspond to individual customers
in FIGS. 1, 2, and 3 respectively.
[0050] FIG. 4 shows: a first customer location with ReachDSL modem
4111 and analog POTS phone(s) 4113 connected to a segment of a
first subscriber loop with at most one load coil 4151; a second
customer location with ReachDSL modem 4211 and analog POTS phone(s)
4213 connected to a segment of a second subscriber loop with at
most one load coil 4251; and a third customer location with
ReachDSL modem 4311 and analog POTS phone(s) 4313 connected to a
segment of a third subscriber loop with at most one load coil 4351.
The three segments of the subscriber loops are further connected to
conversion equipment with multiplexing/inverse multiplexing
(mux/imux) 4550. In FIG. 4, ReachDSL technology is used over each
of the subscriber loops with load coils 4151, 4251, and 4351
between conversion equipment with mux/imux 4550 and ReachDSL modems
4111, 4211, and 4311. As previously mentioned, Paradyne's ReachDSL
technology is capable under certain conditions of working over
local loops or segments of local loops containing one load coil.
Furthermore, conversion equipment with mux/imux 4550 is connected
to central office equipment with multiplexing/inverse multiplexing
(mux/imux) 4560 over one or more segments of loaded loops. In the
preferred embodiments of the present invention DSL service over the
loop segments between conversion equipment 4550 and central office
equipment 4560 would be provided using the techniques described in
the patent application with attorney docket number 61607-1780,
entitled "Digital Subscriber Line Service Over Loaded Loops."
[0051] In FIG. 4, conversion equipment with mux/imux 4550 is
connected to CO equipment with mux/imux 4560 over a first segment
of a subscriber loop with at least one load coil (arbitrarily shown
as load coils 4651, 4653, and 4655), over a second segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 4751, 4753, and 4755), and over a third segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 4851, 4853, and 4855). In addition, the central office
equipment 4560 is further connected to PSTN 4950 and data network
4960, which in the non-limiting preferred embodiments of the
present invention generally provide circuit-switching and
packet-switching respectively.
[0052] The multiplexing and inverse multiplexing in conversion
equipment with mux/imux 4550 and central office equipment with
mux/imux 4560 means that the number of loaded loop segments between
devices 4550 and 4560 do not have to match the number of loaded
loop segments going from the conversion equipment 4550 to each
customer location or customer premises. Instead, the segments of
loaded loops between conversion equipment 4550 and central office
equipment 4560 can be shared to support the plurality of
subscribers. Such a configuration allows efficient usage of
bandwidth over the loaded loop segments between conversion
equipment 4550 and CO equipment 4560. Furthermore, such
multiplexing and inverse multiplexing between conversion equipment
4550 and CO equipment 4560 on the wiring pairs, which previously
may or may not have been used as segments of active subscriber
loops before installation of the conversion equipment 4550, allows
efficiency advantages based on the fact that most subscribers do
not all try to access POTS and/or DSL service simultaneously. Thus,
the multiplexing and inverse multiplexing can be designed with
various contention ratios as subscribers contend for bandwidth
access. One skilled in the art will be aware that the PSTN and data
networks generally also are designed using contention as a way to
increase network efficiency based on statistical profiles of
requests for service generally becoming more predictable as the
number of users in a contention group increases (i.e., the variance
generally decreases as the number of samples increases).
[0053] FIG. 5 shows: a first customer location with ADSL modem 5111
and analog POTS phone(s) 5113 connected to a segment of a first
subscriber loop without any load coils; a second customer location
with ADSL modem 5211 and analog POTS phone(s) 5213 connected to a
segment of a second subscriber loop without any load coils; and a
third customer location with ADSL modem 5311 and analog POTS
phone(s) 5313 connected to a segment of a third subscriber loop
without any load coils. The three segments of the subscriber loops
are further connected to conversion equipment with
multiplexing/inverse multiplexing (mux/imux) 5550. In FIG. 5, ADSL
technology is used over each of the three subscriber loops (or loop
segments) between conversion equipment with mux/imux 5550 and ADSL
modems 5111, 5211, and 5311. As previously mentioned, unlike
ReachDSL modems, standard ADSL modems are not capable of working
over local loops or segments of local loops containing even one
load coil. Furthermore, conversion equipment with mux/imux 5550 is
connected to central office equipment with multiplexing/inverse
multiplexing (mux/imux) 5560 over one or more segments of loaded
loops. In the preferred embodiments of the present invention DSL
service over the loop segments between conversion equipment 5550
and central office equipment 5560 would be provided using the
techniques described in the patent application with attorney docket
number 61607-1780, entitled "Digital Subscriber Line Service Over
Loaded Loops."
[0054] In FIG. 5, conversion equipment with mux/imux 5550 is
connected to CO equipment with mux/imux 5560 over a first segment
of a subscriber loop with at least one load coil (arbitrarily shown
as load coils 5651, 5653, and 5655), over a second segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 5751, 5753, and 5755), and over a third segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 5851, 5853, and 5855). In addition, the central office
equipment 5560 is further connected to PSTN 5950 and data network
5960, which in the non-limiting preferred embodiments of the
present invention generally provide circuit-switching and
packet-switching respectively.
[0055] The multiplexing and inverse multiplexing in conversion
equipment with mux/imux 5550 and central office equipment with
mux/imux 5560 means that the number of loaded loop segments between
devices 5550 and 5560 do not have to match the number of loaded
loop segments going from the conversion equipment 5550 to each
customer location or customer premises. Instead the segments of
loaded loops between conversion equipment 5550 and central office
equipment 5560 can be shared to support the plurality of
subscribers. Such a configuration allows efficient usage of
bandwidth over the loaded loop segments between conversion
equipment 5550 and CO equipment 5560. Furthermore, such
multiplexing and inversion multiplexing between conversion
equipment 5550 and CO equipment 5560 on the wiring pairs, which
previously may or may not have been used as segments of active
subscriber loops before installation of the conversion equipment
5550, allows efficiency advantages based on the fact that most
subscribers do not all try to access POTS and/or DSL service
simultaneously. Thus, the multiplexing and inverse multiplexing can
be designed with various contention ratios as subscribers contend
for bandwidth access. One skilled in the art will be aware that the
PSTN and data networks generally also are designed using contention
as a way to increase network efficiency based on statistical
profiles of requests for service generally becoming more
predictable as the number of users in a contention group increases
(i.e., the variance generally decreases as the number of samples
increases).
[0056] FIG. 6 is similar to FIG. 4, but unlike FIG. 4 the three
customer locations in FIG. 6 do not have analog POTS service. FIG.
6 shows: a first customer location with ReachDSL modem 6111
connected to a segment of a first subscriber loop with at most one
load coil 6151; a second customer location with ReachDSL modem 6111
connected to a segment of a second subscriber loop with at most one
load coil 6251; and a third customer location with ReachDSL modem
6311 connected to a segment of a third subscriber loop with at most
one load coil 6351. The three segments of the subscriber loops are
further connected to conversion equipment with multiplexing/inverse
multiplexing (mux/imux) 6550. In FIG. 6, ReachDSL technology is
used over each of the subscriber loops with load coils 6151, 6251,
and 6351 between conversion equipment with mux/imux 6550 and
ReachDSL modems 6111, 6211, and 6311. As previously mentioned,
Paradyne's ReachDSL technology is capable under certain conditions
of working over local loops or segments of local loops containing
one load coil. Furthermore, conversion equipment with mux/imux 6550
is connected to central office equipment with multiplexing/inverse
multiplexing (mux/imux) 6560 over one or more segments of loaded
loops. In the preferred embodiments of the present invention DSL
service over the loop segments between conversion equipment 6550
and central office equipment 6560 would be provided using the
techniques described in the patent application with attorney docket
number 61607-1780, entitled "Digital Subscriber Line Service Over
Loaded Loops."
[0057] In FIG. 6, conversion equipment with mux/imux 6550 is
connected to CO equipment with mux/imux 6560 over a first segment
of a subscriber loop with at least one load coil (arbitrarily shown
as load coils 6651, 6653, and 6655), over a second segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 6751, 6753, and 6755), and over a third segment of a
subscriber loop with at least one load coil (arbitrarily shown as
load coils 6851, 6853, and 6855). In addition, the central office
equipment 4560 is further connected to PSTN 6950 and data network
6960, which in the non-limiting preferred embodiments of the
present invention generally provide circuit-switching and
packet-switching respectively.
[0058] The multiplexing and inverse multiplexing in conversion
equipment with mux/imux 6550 and central office equipment with
mux/imux 6560 means that the number of loaded loop segments between
devices 6550 and 6560 do not have to match the number of loaded
loop segments going from the conversion equipment 6550 to each
customer location or customer premises. Instead the segments of
loaded loops between conversion equipment 6550 and central office
equipment 6560 can be shared to support the plurality of
subscribers. Such a configuration allows efficient usage of
bandwidth over the loaded loop segments between conversion
equipment 6550 and CO equipment 6560. Furthermore, such
multiplexing and inversion multiplexing between conversion
equipment 6550 and CO equipment 6560 on the wiring pairs, which
previously may or may not have been used as segments of active
subscriber loops before installation of the conversion equipment
6550, allows efficiency advantages based on the fact that most
subscribers do not all try to access POTS and/or DSL service
simultaneously. Thus, the multiplexing and inverse multiplexing can
be designed with various contention ratios as subscribers contend
for bandwidth access. One skilled in the art will be aware that the
PSTN and data networks generally also are designed using contention
as a way to increase network efficiency based on statistical
profiles of requests for service generally becoming more
predictable as the number of users in a contention group increases
(i.e., the variance generally decreases as the number of samples
increases).
[0059] Although FIGS. 4, 5, and 6 show all customer premises to be
configured exactly the same, one skilled in the art will be aware
that equipment can be built supporting multiple configurations.
Thus, a single piece of conversion equipment with mux/imux may
support different customer configurations such as but not limited
to, ReachDSL plus POTS on a loop (or loop segment) with one load
coil, ReachDSL plus POTS on a loop with no load coils, ADSL plus
POTS on a loop with no load coils, ADSL without POTS on a loop with
no load coils, ReachDSL without POTS on a loop with one load coil,
and ReachDSL without POTS on a loop with no load coils.
[0060] Furthermore, one skilled in the art will be aware of various
multiplexing techniques that usually are performed to separate and
combine various information flows to share common facilities or
resources for propagation of the electromagnetic signals while
still allowing a receiver to pick out the proper signal. As
information flows generally are carried in electromagnetic waves
due to the propagation speed of the waves, the different
characteristics of electromagnetic waves can be used to develop
various multiplexing techniques. For example, some common
parameters of electromagnetic waves are time, frequency, space,
direction of propagation, and polarity that each can be used to
separate electromagnetic signals sharing some common resource. Some
non-limiting examples of the ways these different characteristics
of electromagnetic waves lead to multiplexing techniques include,
but are not limited to, time-division multiplexing (TDM) for
sharing a common communications medium, frequency division
multiplexing (FDM) for sharing a common communications medium, and
spatial or distance separation of electromagnetic signals to
prevent interference. Furthermore, given that the wavelength and
the frequency of an electromagnetic wave are related to the speed
of light, wavelength-division multiplexing (WDM) is a form of FDM.
In addition, coding theory provides another way for separating
information that is used in code-division multiple access (CDMA)
technologies. In addition, the two common types of TDM are fixed or
static TDM, with the 56/64 kbps DSO channels in the PSTN being a
common example, and statistical or dynamic TDM, with packet
networks being a common example.
[0061] While various multiplexing techniques could be used over
loaded loops (or loaded loop segments) between conversion equipment
4550 and CO equipment 4560 in FIG. 4, between conversion equipment
5550 and CO equipment 5560 in FIG. 5, and between conversion
equipment 6550 and CO equipment 6560 in FIG. 6, some multiplexing
techniques are more cost efficient than others given various
amounts of information that needs to be communicated between two
points and the bandwidth resources available to provide for the
communication. Often changes in the costs of the electronic devices
used to implement various types of multiplexing can change the most
cost efficient type of multiplexing to use in a product. However,
time-division multiplexing certainly is one multiplexing technique
that has been commonly used in low cost devices because the digital
devices for TDM generally have followed Gordon Moore's law of
increasing in performance and decreasing in price. Although some of
the older analog trunk equipment in the phone network used FDM and
some point-to-point microwave toll hops used FDM, the PSTN
generally moved towards TDM with the introduction of digital
technologies such as T-carrier (and later technologies such as
SONET). If not for the problems of powering an ISDN terminal
adapter (TA) in emergency situations, TDM would have been used for
more DSL deployments to customer premises. ADSL was designed to be
frequency-division multiplexed above the POTS baseband because of
the lifeline emergency requirements for the 0-4 KHz powered POTS
interface.
[0062] Generally, service providers are required to use equipment
that meets certain reliability standards such as NEBS (Network
Equipment/Building System) compliance in various locations in their
networks. Thus, service providers are familiar with designing their
networks and network powering to meet various reliability
requirements. Unlike requiring customers to design their own
powering reliability that was one of the pitfalls of using BRI ISDN
for POTS replacement, service providers are in the business of
providing reliable power to their network equipment. Because
service providers can design their network to deliver high enough
reliability power to equipment located at a wiring concentration
center where the conversion equipment 4550, 5550, or 6550 may be
located, the multiplexing across the loops or loop segments between
the conversion equipment 4550, 5550, or 6550 and the CO equipment
4560, 5560, or 6560 no longer would necessarily have to carry a
native 0-4 KHz POTS interface that is frequency-division
multiplexed on each loop. Instead, one skilled in the art will be
aware that the POTS service can be digitized to be carried in a
fixed TDM manner of 56 kbps or 64 kbps A-law or .mu.-law speech and
associated digital signaling messages. Also, one skilled in the art
will be aware that various fixed and variable format improved voice
compression techniques have been developed since 56/64 kbps A-law
or .mu.-law speech was developed for T1/E1. Also, one skilled in
the art will be aware that various packetized voice techniques have
been developed for communicating voice and voice-call signaling
over the statistical time-division multiplexing of packet
networks.
[0063] While the efficiency improvements from multiplexing and
aggregation are known to one of ordinary skill in the art, other
factors also are important in communications systems. For example,
aggregating a large amount of traffic through various concentration
techniques such as multiplexing generally can improve efficiency,
but may alternatively hurt network reliability as the concentrated
locations become single points of failure affecting the
communications traffic of a larger number of users. Thus, in
addition to efficiency, redundancy is another important criteria in
network design.
[0064] While concentrating the data and voice traffic of many
customers into a single multiplexed digital stream offers
efficiency improvements, it may create some more concentrated
potential points of failure. Because basic native 0-4 KHz POTS
service is generally considered a lifeline service demanding high
reliability, there may be some justifications for not digitizing
and multiplexing all the native 0-4 KHz POTS communication from
subscribers. The embodiments of the present invention also will
work if the native POTS communication is not digitized, but is
instead carried back to a corresponding POTS line card in a CO or
DLC for each of the subscriber loops with both native POTS and DSL
service that are connected to the conversion equipment 5550, 6550,
or 7550. Generally, the DSL data service provided from the
conversion equipment 5550, 6550, or 7550 has lower reliability
requirements than lifeline POTS service. Thus, the DSL data traffic
from multiple customers generally would be multiplexed together and
inverse multiplexed over multiple loops, even though a native 0-4
KHz POTS service for a customer might not be multiplexed in with
this DSL data.
[0065] One skilled in the art will be aware that there often are
various trade-offs with different multiplexing strategies that
affect criteria such as, but not limited to, efficiency,
reliability, and/or security. Generally, the DSL data service and
the native 0-4 KHz POTS service have different reliability criteria
for most customer situations. However, the reliability criteria
certainly can vary from customer to customer. Thus, embodiments of
the present invention certainly could be configured to allow
different customer-by-customer policy decisions for the
multiplexing techniques (or lack thereof) to handle a particular
customer's voice and/or data communication, while meeting the
various performance criteria demanded by the customer.
[0066] As a non-limiting example, a customer premises such as a
fire house, police precinct, or other emergency service location
may need even higher reliability native 0-4 KHz POTS service than a
normal residential customer premises. As a non-limiting example,
the preferred embodiments of the present invention can be
configured to utilize various multiplexing techniques to carry the
DSL data and native 0-4 KHz POTS "lifeline" interfaces of various
residential customers when sufficiently-capable powering
reliability to the conversion equipment exists to meet the
reliability requirements of using a digitized POTS service to carry
the residential customer's POTS lifeline connectivity. However, the
preferred embodiments of the present invention may provide DSL data
service to a firehouse on the same loop providing a native 0-4 KHz
POTS interface to the firehouse, with the firehouse requiring its
POTS service to meet even higher reliability requirements than
normal residential POTS. To the extent that these higher
reliability requirements of the firehouse are not met by the
powering reliability and redundancy network design choices for
supplying the conversion equipment 5550, 6550, or 7550 with power,
the native 0-4 KHz POTS interface for just the firehouse customer
premises may be carried back to a line card in a CO or DLC without
being digitized and/or multiplexed by the conversion equipment
5550, 6550, or 7550. thus, the POTS service to the firehouse
customer may be wired different than the POTS service to other POTS
residential customers though the same conversion equipment 5550,
6550, or 7550 may provide loaded loop DSL service to both.
[0067] Thus, the embodiments of the present invention also will
work if only some or none of the native POTS communication is
digitized. Instead, for any particular chosen customer loop with
native POTS service, the POTS service could be continued to be
carried back from conversion equipment 5550, 6550, or 7550 to a
POTS line card in a CO or DLC on its local loop that just carries
POTS service without utilizing any multiplexing efficiencies of
concentration. While such a configuration may be inefficient in
bandwidth usage, it may offer additional reliability and redundancy
advantages that are important for some lifeline POTS
situations.
[0068] FIG. 7 shows more detail of conversion equipment to support
a single subscriber loop. In FIG. 7, a customer premises 7001
comprises a ReachDSL CPE-side modem 7011, a POTS phone 7013 (which
may be other types of equipment with a customer-side POTS
interface), and a microfilter (MF) 7015. One skilled in the art
will be familiar with DSL microfilters. The customer premises 7001
is connected to conversion equipment 7550 over a subscriber loop
(or a segment of a subscriber loop), which may have up to one load
coil 7051 over which ReachDSL will operate. Conversion equipment
7550 further comprises ReachDSL CO-side modem 7061 and a POTS
splitter (PS) 7063, whose function will be known to one of ordinary
skill in the art of DSL technology. Furthermore, the POTS splitter
7063 is connected to a POTS subscriber line interface card (SLIC)
7065 that generally would implement the functions of a CO-side
standard POTS interface in the same way that a line card in a CO
switch or DLC provide such a POTS interface. The ReachDSL CO modem
7061 and POTS SLIC 7065 generally would present a digital interface
to multiplexer 7555. The digital interface of POTS SLIC 7065 could
be the standard 56/64 kbps DSO PCM (Pulse Code Modulation) voice,
some other fixed bandwidth compressed voice format such as but not
limited to 32 kbps ADPCM (Adaptive Differential Pulse Code
Modulation), some variable bandwidth compressed voice format such
as but not limited to CELP (Code Excited Linear Prediction) voice,
or some packetized digital voice format.
[0069] The digital information from ReachDSL CO modem 7061 and POTS
SLIC 7065 is multiplexed together in multiplexer/inverse
multiplexer (mux/imux) 7555 before being passed on to loaded loop
transceiver 7650. The conversion equipment 7550 is connected to
CO-side equipment 7560 over one or more subscriber loops (or
subscriber loop segments). Active subscriber loops between the
conversion equipment 7550 and CO-side equipment 7560 have a
conversion-equipment-side loaded loop transceiver (LLT) 7650 in
communication with a CO-side loaded loop transceiver (LLT) 7660.
The CO-side loaded loop transceiver (LLT) 7660 is connected to data
and POTS interface 7955 that in the preferred embodiment might
separate out the data from the ReachDSL modem for connection to a
data network, while separating out the customers POTS information
for connection to the PSTN. Potentially, the POTS information might
not be converted back to individual analog POTS loops but could be
provided to a central office switch in a GR-303 digital format that
is commonly used for interfacing CO switches to digital loop
carrier (DLC) systems, which use 56/64 kbps PCM voice and TDM for
digitally carrying the information from many POTS subscriber lines
to a CO switch in the PSTN. Alternatively, many telephone companies
already have standard packetized voice interfaces to the PSTN. One
such interface is the broadband loop emulation service (BLES) that
generally is based on VoDSL using ATM Adaptation Layer 2 (AAL2).
The packetized voice of the BLES interface might use 32 kbps ADPCM
voice encoding.
[0070] In general, to support a native 0-4 KHz POTS interface on
the loop going to the subscriber premises, the choice of a
particular type of voice encoding and whether statistical TDM
(i.e., packets) or fixed TDM is used to carry digitized voice over
the loaded loops is independent from the choices used in
interfacing voice to the PSTN. One voice format generally can be
converted to another voice format. However, such format conversions
generally require substantial processing power, which may be a
reasonable tradeoff in exchange for a more efficient voice encoding
and multiplexing scheme that reduces bandwidth demands on the
loaded loops between the conversion equipment 7550 and CO-side
equipment 7560. A 11 possible types of digitized voice encodings
and/or multiplexing schemes for carrying voice the conversion
equipment 7550 and CO-side equipment 7560 to support native POTS
service in the 0-4 KHz bandwidth are intended to be within the
scope of the present invention. Also, all possible formats for
interfacing voice to the PSTN are intended to be within the scope
of the present invention.
[0071] FIG. 8 shows the connections for a multiple
subscriber/multiple customer premises implementation of the
configuration of FIG. 7. Generally, each component acts similarly
to the way it acted in FIG. 7. The first customer premises 8101
contains ReachDSL CPE modem 8111, POTS phone 8113, and microfilter
(MF) 8115 and is connected over a subscriber loop with up to one
load coil 8151 to conversion equipment 8550. Conversion equipment
contains ReachDSL CO modem 8161, POTS splitter (PS) 8163, and POTS
subscriber line interface card (SLIC) 8165 to support the data and
POTS of the first customer premises 8101. The second customer
premises 8201 contains ReachDSL CPE modem 8211, POTS phone 8213,
and microfilter (MF) 8215 and is connected over a subscriber loop
with up to one load coil 8251 to conversion equipment 8550.
Conversion equipment contains ReachDSL CO modem 8261, POTS splitter
(PS) 8263, and POTS subscriber line interface card (SLIC) 8265 to
support the data and POTS of the second customer premises 8201. The
third customer premises 8301 contains ReachDSL CPE modem 8311, POTS
phone 8313, and microfilter (MF) 8315 and is connected over a
subscriber loop with up to one load coil 8351 to conversion
equipment 8550. Conversion equipment contains ReachDSL CO modem
8361, POTS splitter (PS) 8363, and POTS subscriber line interface
card (SLIC) 8365 to support the data and POTS of the third customer
premises 8301. The digital information flows for supporting POTS
and data at each of the three customer premises 8101, 8201, and
8301 are multiplexed together and inverse multiplexed across at
least one and probably a plurality of loops between conversion
equipment 8550 and CO-side equipment 8560. Each of the loops has a
conversion equipment-side loaded loop transceiver (LLT) and a
CO-side loaded loop transceiver (LLT). For the first loop 8600
between conversion equipment 8550 and CO-side equipment 8560, LLT
8650 is connected over a loaded loop 8600 with at least one load
coil (arbitrarily shown as load coils 8651, 8653, and 8655) to LLT
8660. For the second loop 8700 between conversion equipment 8550
and CO-side equipment 8560, LLT 8750 is connected over a loaded
loop 8700 with at least one load coil (arbitrarily shown as load
coils 8751, 8753, and 8755) to LLT 8760. On the CO-side, the
information flows are properly multiplexed, demultiplexed, and/or
inverse multiplexed before being passed to the proper networks
through data and POTS interface 8955. As described previously with
respect to FIG. 7, although individual POTS loops could be used for
data and POTS interface 8955, the higher concentrations of POTS
interfaces at this point make it preferred to use some form of
multiplexed POTS interface to the PSTN such as, but not limited to,
the GR-303 interface that is used for DLCs. Alternatively, many
telephone companies already have standard packetized voice
interfaces to the PSTN. One such interface is the broadband loop
emulation service (BLES) that generally is based on VoDSL using ATM
Adaptation Layer 2 (AAL2). The packetized voice of the BLES
interface might use 32 kbps ADPCM voice encoding.
[0072] In general, to support a native 0-4 KHz POTS interface on
the loop going to the subscriber premises, the choice of a
particular type of voice encoding and whether statistical TDM
(i.e., packets) or fixed TDM is used to carry digitized voice over
the loaded loops is independent from the choices used in
interfacing voice to the PSTN. One voice format generally can be
converted to another voice format. However, such format conversions
generally require processing horsepower, which may be a reasonable
tradeoff in exchange for a more efficient voice encoding and
multiplexing scheme that reduces bandwidth demands on the loaded
loops between the conversion equipment 8550 and CO-side equipment
8560. All possible types of digitized voice encodings and/or
multiplexing schemes for carrying voice the conversion equipment
8550 and CO-side equipment 8560 to support native POTS service in
the 0-4 KHz bandwidth are intended to be within the scope of the
present invention. Also, all possible formats for interfacing voice
to the PSTN are intended to be within the scope of the present
invention.
[0073] As can be seen from FIG. 8, the multiplexing/inverse
multiplexing leads to the number (3) of loops (or loop segments)
between the conversion equipment 8550 and the three customer
premises 8101, 8201, and 8301 being potentially different from the
number (2) of loops (or loop segments) between LLTs 8650/8670 and
LLTs 8750/8760. Various criteria, such as but not limited to, the
amount of bandwidth available on a loaded loop between conversion
equipment 8550 and CO-side equipment 8560, the expected and/or peak
data bandwidth utilization by expected customers, the techniques
used for encoding and/or compressing voice, the allowable
contention ratios for customers, as well as many other factors
would go into the capacity planning to properly size a
configuration with a reasonable number of loaded loops between
conversion equipment 8550 and CO-side equipment 8560 to meet
various quality of service (QoS) goals in a contract between a
service provider and a customer.
[0074] In addition, various types of techniques can be used for
carrying the bi-directional communications between conversion
equipment 8550 and CO-side equipment 8560. To the extent that the
loaded loops between conversion equipment 8550 and CO-side
equipment no longer carry a native analog POTS interface in the 0-4
KHz frequency band, this bandwidth is freed up for digital
communication. Also, the potentially large number of loops between
conversion equipment 8550 and CO-side equipment 8560 allows for
some additional duplexing strategies that were not as practical for
DSL delivery to a customer premises that generally is initially
wired with only two loops under the standard operating line
installation procedures of telcos. With a large number of loaded
loops between conversion equipment 8550 and 8560, four-wire
duplexing is one duplexing strategy that removes all the
frequency-dependent problems of frequency-division duplexing (FDD)
and the frequency-dependent effects to the echo cancellation noise
floor in pure echo cancelled duplex (ECD). As was discussed in more
detail in the patent application with attorney docket number
61607-1780, entitled "Digital Subscriber Line Service Over Loaded
Loops", and filed the same day that was previously incorporated by
reference in its entirety herein, load coils on telco loops
introduce frequency-dependent problems. However, four-wire
duplexing and time-division duplexing (TDD) and/or adaptive
time-division duplexing (ATDD) generally are not affected by
frequency-dependent impediments such as load coils.
[0075] One skilled in the art will be aware that various types of
communication applications have different directional traffic
patterns. For instance, telephone conversations generally are
symmetric, while internet access where a user downloads many web
pages generally is asymmetric. Asymmetric Digital Subscriber Line
(ADSL) was designed given these traffic patterns that normally
occur as subscribers use DSL lines for internet access. Using
four-wire duplexing techniques on more than two loops allows a
service provider to allocate asymmetric amounts of bandwidth for
each direction of communication over the loaded loops. In general,
when only four wires (or two pair) are available, four-wire
duplexing uses one wire pair for one direction of communication and
the other pair form the opposite direction of communication.
Basically, each pair is used in a simplex fashion to just support a
single direction of communication. Similarly, multiple loops
between conversion equipment 8550 and CO-side equipment 8560 could
be configured in a simplex fashion to support each direction of
communication. As a non-limiting example, suppose there are ten
loaded loops (with all the same bit rate capacities) between
conversion equipment 8550 and CO-side equipment 8560. Further
suppose that data traffic patterns whether determined statically or
dynamically based on data demand have an asymmetric ratio
suggesting 70% of the traffic is downstream from the network to the
user, while 30% of the traffic is upstream from the user to the
network. In such a non-limiting example, seven of the loops between
conversion equipment 8550 and CO-side equipment 8560 could be used
to support simplex downstream communication, while three of the
loops between conversion equipment 8550 and CO-side equipment 8560
could be used to support simplex upstream communication. Obviously
an equal number of loops could be used in simplex for each
direction of traffic if the traffic patterns are more symmetric.
Also, TDD/ATDD might be used to subdivide the direction of
communication for one or a few loops not using simplex
communications to generally match any particular requirement ratios
of upstream to downstream bandwidth.
[0076] In addition to a fixed asymmetrical allocation of some wire
pairs to one direction of communication and some wire pairs to
another direction of communication, the asymmetry/symmetry of the
use of various pairs of wires of loaded loops can be varied
dynamically and statistically based on changing data demands for
each direction of communication. For example, with four loaded
loops between conversion equipment 8550 and CO-side equipment 8560
that each only are used in a simplex fashion, bi-directional
communications can be continuously allowed while the
asymmetry/symmetry varies from three pairs in one direction and one
pair in the other direction to two pairs in one direction and two
pairs in the other direction before varying again to one pair in
one direction and three pairs in the other direction. Thus, these
assignments of a wire pair or loop to carry a particular direction
of traffic can be static or dynamic, and the dynamic allocations
may be automatic and/or adaptive to network conditions such as, but
not limited to, the demands for bandwidth in a particular direction
balanced against the demands for bandwidth in the opposite
direction.
[0077] Furthermore, other potential solutions to bi-directional
communication are possible. Certainly, multiple wire pairs can each
provide duplex communications using various techniques such as, but
not limited to, TDD/ATDD, pure ECD, and Extended Performance ECD
that is described in U.S. patent application Ser. No. 10/420,204,
which is entitled "Extended-Performance Echo-Cancelled Duplex (EP
ECD) Communication", was filed on Apr. 22, 2003, and is
incorporated by reference in its entirety herein. Also,
asymmetrical rate echo cancellation can also be used in the
duplexing.
[0078] With regard to the inverse multiplexing of various digital
bit streams of both data (potentially including synthesized or
derived voice from technologies such as but not limited to VoDSL,
VoIP, and VoATM provided in the DSL channel on a subscriber loop)
and voice (from the native POTS channel on a subscriber line),
various techniques for inverse multiplexing information flows over
multiple links are known in the art. Some non-limiting examples
include the multi-link point-to-point protocol (MLPPP), the
ethernet link aggregation protocol, and the open shortest path
first (OSPF) routing protocol. However, these techniques are not
necessarily well suited to the preferred embodiments of the present
invention. In general, the listed example techniques all are
designed for large data frames and often introduce latency that is
detrimental to real time applications such as carrying POTS voice.
Certainly one or more loaded loops between conversion equipment
8550 and CO-side equipment 8560 could utilized fixed TDM time slots
to carry the POTS voice with the time slots being filled in a round
robin fashion in the same way T1 carries digitized POTS phone
calls. However, such a solution might not be the most
efficient.
[0079] Instead various statistical time-division multiplexing
(STDM) techniques offer the advantage of using any and potentially
all of the loaded loops to carry either data from/to DSL modems as
well as the digitized voice from/to analog POTS phones. A small
packet size in the statistical multiplexing can be used to help
reduce latency. Thus, ATM would seem to be a preferable method for
handling the multiplexing/inverse multiplexing over the loaded
loops between conversion equipment 8550 and CO-side equipment 8560.
ATM offers the bandwidth utilization efficiencies of statistical
multiplexing, while addressing the latency issues with a small cell
size of 53 octets including 48 octets of data and a 5 octet header.
In addition, ATM has some well-developed quality of service (QoS)
mechanisms that can be used to meet the differing performance
requirements of both real-time voice and computer data. In exchange
for its advantages, ATM introduces the minor penalty of a larger
amount of the bandwidth being used for communicating header
information than would be used in larger size packets in frame
relay or IP.
[0080] The inverse multiplexing for MLPPP, ethernet link
aggregation, and ATM in Inverse Multiplexing over ATM (IMA)
generally are OSI (Open Systems Interconnect) layer two constructs,
while the inverse multiplexing of IP datagrams across multiple OSPF
links generally is an OSI layer three construct. One skilled in the
art will be aware that inverse multiplexing can also be performed
at the OSI level one physical layer. The bandwidth on demand
interoperability group (BONDING) developed a physical layer inverse
multiplexing standard for 56/64 kbps DSOs that was primarily used
for digital video. Various physical layer inverse multiplexing
techniques also could be utilized in the preferred embodiments of
the present invention to interleave bits at the physical medium
dependent (PMD) sublayer.
[0081] Also, when dealing with inverse multiplexing of multiple
links, the propagation delay time over each link can sometimes
vary. For example, on a 128 kbps circuit-switched video call over
two ISDN B-channels, each DSO phone call could follow different
paths through the PSTN such that one DSO is routed over a land line
and the other DSO is routed over a satellite channel. The resulting
potentially large delay differences generally should be addressed
by the inverse multiplexing technology. However, in the preferred
embodiments of the present invention, the lengths of the loops
between conversion equipment 8550 and CO-side equipment 8560 will
be approximately the same. As a result, electromagnetic signals
will have approximately the same propagation time over the
loops.
[0082] However, there are issues in inverse multiplexing when the
data rates of the loops are different. For example, inverse
multiplexing information across two loops with one loop running at
19.2 kbps and another at 64 kbps presents problems. One
non-limiting solution to this problem would be to assume that all
loops will be adjusted to run at the speed of the lowest currently
operating loop. Then a determination of the anticipated inverse
multiplexing throughput is made. Next, drop out the lowest speed
loops and redo the calculations. Once no loops remain, then the
throughput for all the possible inverse multiplexing configurations
with the loops running at the same data rate will be determined.
Choosing the highest throughput configuration will tend to maximize
performance, and some of the lowest data rate loops may well be
dropped from the inverse multiplexing group, while some of the
highest data rate loops may well have their data rate downgraded.
With similar propagation delays and the same data rate on each loop
in the multiplexed group, a simple round-robin inverse multiplexing
scheme could be employed in assigning ATM cells to queues
associated with particular loaded loops in the preferred
embodiments of the present invention.
[0083] Various techniques can be used for physical layer inverse
multiplexing and some non-limiting examples are discussed. One
potential non-limiting physical layer inverse multiplexing
technique might perform some of the following functions. First,
equipment could measure the differential propagation delays across
multiple modem receivers. Then, received data could be buffered to
compensate for the differential delays. One non-limiting technique
for measuring the delay might be to introduce an overhead channel
that results in a reduction in overall throughput. Some potential
problems with such an inverse multiplexing strategy might include
lost capacity, detection of loss of one or more wire pairs, and/or
a synchronization delay while waiting for the long training
sequences of modems to complete.
[0084] Some patents and patent applications that are relevant to
providing physical layer inverse multiplexing include: U.S. Pat.
No. 4,630,286 to William L. Betts, entitled "Device for
Synchronization of Multiple Telephone Circuits", filed on Oct. 10,
1984, and issued on Dec. 16, 1986, which is incorporated by
reference in its entirety herein; U.S. Pat. No. 4,637,035 to
William L. Betts, entitled "Digital Modem for Multiple Telephone
Circuits", filed on Feb. 16, 1984, and issued on Jan. 13, 1987,
which is incorporated by reference in its entirety herein; U.S.
Pat. No. 4,734,920 to William L. Betts, entitled "High Speed Modem
for Multiple Communication Circuits", filed on Oct. 10, 1984, and
issued on Mar. 29, 1988, which is incorporated by reference in its
entirety herein; U.S. Pat. No. 5,134,633 to Jean-Jacques Werner,
entitled "Digital Communications Synchronization Scheme", filed on
Nov. 30, 1990, and issued on Jul. 28, 1992, which is incorporated
by reference in its entirety herein; U.S. Pat. No. 5,163,066 to
Robert L. Cupo and Cecil W. Farrow, entitled "Synchronizing the
Operation of Multiple Equalizers in a Digital Communications
System", filed on Mar. 24, 1991, and issued on Nov. 10, 1992, which
is incorporated by reference in its entirety herein; and U.S.
patent application Ser. No. 09/534,696, applied for by William L.
Betts, entitled "Space Diversity Trellis Interleaver System and
Method", and filed on Mar. 24, 2000, which is incorporated by
reference in its entirety herein.
[0085] In general, U.S. Pat. No. 4,630,286 to Betts uses an
out-of-band phase-shift synchronization signal to detect
differential delay without the losses of data rate caused by
overhead. Also, U.S. Pat. No. 4,637,035 to Betts generally
identifies the use of a high-speed signal processor to handle
multiple lower speed channels. In addition, U.S. Pat. No. 4,734,920
to Betts generally identifies a full multi-pair system using a
single processor and measuring differential delay across the
multiple pairs using the modem training sequence. Furthermore, U.S.
Pat. No. 5,163,066 to Cupo et al. generally describes multi-pair
equalizers, while U.S. Pat. No. 5,134,633 to Werner generally
describes differential delay synchronization for a plurality of
channels.
[0086] Furthermore, U.S. patent application Ser. No. 09/534,696 of
Betts describes several methods to allow independent data rates or
constellation densities on each pair or loop. The method of a
single processor handling multiple pairs has at least the
advantages of space diversity, lower throughput delay, and reduced
complexity, which are important in low symbol rate transmissions
that would likely be used on loaded loops.
[0087] Moreover, the space diversity interleaving, which is
described in the U.S. patent application Ser. No. 09/534,696 of
Betts, can be incorporated within the multiplexing/inverse
multiplexing unit, 8555. Multiple LLTs 8650 and 8750 in conversion
equipment 8550 may interleave their transmitted symbols on
alternate time intervals between the pairs 8600 and 8700. This
diversity will improve the performance of a trellis code if used by
the LLT. Alternatively, a single LLT may operate at a higher symbol
rate and transmit alternate symbols on first pair 8600 and then on
second pair 8700. This reduces complexity by using fewer LLTs. It
also reduces latency by operating at a higher symbol rate than
would otherwise be supported on loaded loops. Trellis coding
performance will be improved by alternately transmitting on the
diversity pairs.
[0088] In the preferred embodiments of the present invention
conversion equipment 7550 and 8550 may be located in a remote
terminal (or DLC), in a Service Area Interface (SAI) cabinet, in a
cross-connect cabinet, or in a network interface device (NID) box
that is often mounted on the side of a customer premises.
Obviously, placing the conversion equipment in a NID of a single
dwelling home would likely not allow access to the aggregation and
multiplexing advantages of FIG. 8 over FIG. 7. However, a NID on
the side of a multi-unit apartment building might allow the
multiplexing advantages of FIG. 8. In particular, while the
preferred embodiments of the present invention will function as
intended when placed in many parts of the network various locations
do have advantages and disadvantages, A cross-connect box is a
relatively good place for the conversion equipment because of the
proximity to the customer premises such that the number of loops
with more than one load coil between the conversion equipment and
the customer premises DSL modem such as a ReachDSL modem or an ADSL
modem will be reduced. Thus, placing the conversion equipment in a
cross connect cabinet allows for more DSL service coverage. As a
disadvantage, the cross-connect boxes generally are not very big
and generally do not have a ready source of excess power.
[0089] FIG. 9 shows a potential network configuration, where the
conversion equipment could be placed in a cross-connect box 9550.
In general, the cross-connect box may be connected to many customer
premises (CPs). Each customer premises may have POTS only service,
DSL plus POTS service, or DSL only service. In FIG. 9, customer
premises (CP) 1, 2, 24, 48, 96, 144, 288, and 300 with
corresponding reference numbers 9001, 9002, 9003, 9004, 9005, 9006,
9007, and 9008 are connected over F2 distribution loops 9101 to
cross-connect cabinet 9550. Because two of the F2 distribution
loops 9101 to customer premises 9004 and 9005 each have a single
load coil 9054 and 9055, respectively, standard ADSL generally will
not work over these F2 distribution loops. Instead ReachDSL could
be used to provide DSL service between the cross-connect cabinet
9550 and customer premises 9004 and 9005. The other F2 distribution
loops to customer premises 9001, 9002, 9003, 9006, 9007, and 9008
do not have load coils and could be used to support ADSL or
ReachDSL between the customer premises and conversion equipment
added in the cross-connect cabinet 9550. The cross-connect cabinet
9550 further is connected over F1 feeder loops 9501 to a PSTN
switch 9560 (or line cards in a DLC). The F1 feeder loops 9501,
which were originally part of loaded analog POTS loops to customer
premises, would be converted to digital service using loaded loop
transceivers. In FIG. 9, the F1 feeder loop(s) 9501 have up to
seven load coils 9651, 9652, 9653, 9654, 9655, 9656, and 9657.
[0090] Although the preferred embodiments of the present invention
have primarily been discussed with respect to loaded loops, the
same equipment could operate over unloaded loops (as well as
combinations of loaded and unloaded loops) with potentially even
better performance. Thus, FIG. 10 shows a potential network
configuration, where the conversion equipment could be placed in a
cross-connect box 10550. In FIG. 10, there are no loaded loops. In
general, the cross-connect box may be connected to many customer
premises (CPs). Each customer premises may have POTS only service,
DSL plus POTS service, or DSL only service. In FIG. 10, customer
premises (CP) 1, 2, 24, 48, 96, 144, 288, and 300 with
corresponding reference numbers 10001, 10002, 10003, 10004, 10005,
10006, 10007, and 10008 are connected over F2 distribution loops
10101 to cross-connect cabinet 10550. Because none of the F2
distribution loops 10101 to the customer premises have load coils,
both ReachDSL and standard DSL would work in providing DSL
communication between the customer premises and the conversion
equipment that is installed in cross-connect cabinet 10550. The
cross-connect cabinet 10550 further is connected over F1 feeder
loops 10501 to a PSTN switch 10560 (or line cards in a DLC). The F1
feeder loops 10501, which were originally part of analog POTS loops
to customer premises, would be converted to digital service.
[0091] FIG. 11 shows a more detailed configuration of placing the
conversion equipment in a cross-connect box with various types of
customer premises configurations. In FIG. 11, a first customer
premises with ReachDSL modem 11111 and analog POTS phone(s) 11113
is connected to a subscriber loop (or a segment of a subscriber
loop that did not change when the conversion equipment was
installed). A second customer premises comprises ReachDSL modem and
integrated access device (LAD) 11211 and a POTS phone 11213 that
are directly connected to the subscriber loop with load coil 11251.
POTS phone 11213 utilizes the native 0-4 KHz POTS interface. In
contrast, the communication of phone(s) 11215 are carried in the
DSL channel of the subscriber loop and generally involve digitized
voice encodings that also are packetized for statistical
multiplexing in the preferred embodiments of the present invention.
The ReachDSL modem and IAD 11211 may provide a local POTS interface
such that phone(s) 11215 are standard POTS phones with the IAD
11211 performing the necessary conversion for digital encoding
and/or packetizing. Alternatively, phone(s) 11215 may provide its
own digitized and/or packetized format that is just passed into the
ReachDSL modem 11211.
[0092] In FIG. 11, a third customer premises comprises ReachDSL
modem and integrated access device (IAD) 11311 without any native
POTS service in the 0-4 KHz baseband of the subscriber loop.
Instead, ReachDSL modem and integrated access device (IAD) 11311
offers a derived or synthesized voice service through phone(s)
11315. Often without local backup powering, such a configuration as
the third customer premises loses phone service, when local power
is lost at the customer premises. However, often a user at the
third customer premises has other another lifeline service such as
a cell phone or secondary POTS loop. FIG. 11 also shows a fourth
customer premises with a POTS only service using POTS phone(s)
11413 over the loop with load coil 11451. The fourth customer
premises also could be viewed as the secondary POTS lifeline loop
at the third customer premises.
[0093] All four of the subscriber loops with load coils 11151,
11251, 11351, and 11451 connect to (or through) cross-connect
cabinet 11555. Because the DSL subscriber loops all are shown in
FIG. 11 having one load coil 11151, 11251, and 11351, ReachDSL is
used on these loops instead of standard A DSL, which will not
function even with the single load coil. The DSL subscriber loops
with one load coil 11151, 11251, and 11351 are terminated in
conversion equipment with mux/imux 11550 in cross-connect cabinet
11555. Conversion equipment with mux/imux 11550 is connected over
one or more loaded loops to central office-side equipment with
mux/imux 11560. In FIG. 11, three loaded subscriber loops are
arbitrarily shown between conversion equipment 11550 and CO-side
equipment 11560. Each of these three loaded loops will have a pair
of loaded loop transceivers (LLTs) located on each end of the
loaded loops. Or alternatively, fewer LLTs, each transmitting
symbols on multiple loaded loops to achieve the advantages of
diversity, reduced complexity and reduced latency. FIG. 11
arbitrarily shows the first loaded loop between conversion
equipment 11550 and CO-side equipment 11560 having load coils
11651, 11653, and 11655. In addition, FIG. 11 arbitrarily shows the
second loaded loop between conversion equipment 11550 and CO-side
equipment 11560 having load coils 11751, 11753, and 11755. Also,
FIG. 11 arbitrarily shows the third loaded loop between conversion
equipment 11550 and CO-side equipment 11560 having load coils
11851, 11853, and 11855.
[0094] Furthermore, FIG. 11 shows the basic POTS loop with load
coils 11451, 11951, 11953, and 11955 passing through the
cross-connect box 11555 without going into conversion equipment
11550 before terminating in a line card in PSTN 11950. In general,
the PSTN uses A-law or .mu.-law PCM encoded voice in a fixed 56/64
kbps TDM format. In contrast, the derived or synthesized voice from
phone(s) 11215 or 11315 may pass through data network 11960 using a
different voice encoding than standard PCM and using the
statistical time-division multiplexing (STDM) of packet switching.
Gateway 11965 may be used to covert between different voice
encoding and/or voice packetization formats.
[0095] Referring to FIG. 12, more detail is shown on conversion
equipment 12550 that is arbitrarily connected to three customer
premises. At the first customer premises ReachDSL CPE modem 12111
and POTS phone 12113 are connected to F2 feeder loop 12123, with
the POTS phone 12113 using a microfilter (MF) 12115. ReachDSL CPE
modem 12111 is connected over a loop with up to one load coil 12151
to ReachDSL CO modem 12161. In addition, F2 feeder loop 12123
connects to POTS splitter (PS) 12163, which is further connected to
POTS subscriber line interface card (SLIC) 12165 that provides the
CO-side of the native POTS interface to POTS phone 12113. At the
second customer premises ReachDSL CPE modem 12211 and POTS phone
12213 are connected to F2 feeder loop 12223, with the POTS phone
12213 using a microfilter (MF) 12215. ReachDSL CPE modem 12211 is
connected over a loop with up to one load coil 12251 to ReachDSL CO
modem 12261. In addition, F2 feeder loop 12223 connects to POTS
splitter (PS) 12263, which is further connected to POTS subscriber
line interface card (SLIC) 12265 that provides the CO-side of the
native POTS interface to POTS phone 12213. At the third customer
premises ReachDSL CPE modem 12311 and POTS phone 12313 are
connected to F2 feeder loop 12323, with the POTS phone 12313 using
a microfilter (MF) 12315. ReachDSL CPE modem 12311 is connected
over a loop with up to one load coil 12351 to ReachDSL CO modem
12361. In addition, F2 feeder loop 12323 connects to POTS splitter
(PS) 12363, which is further connected to POTS subscriber line
interface card (SLIC) 12365 that provides the CO-side of the native
POTS interface to POTS phone 12313. One other item to note about
the conversion equipment in FIG. 12, if native 0-4 KHz POTS service
is not provided over to the customer premises, then the POTS
splitters (PS) 12163, 12263, and 12363 as well as POTS SLICs 12165,
12265, and 12365 are not needed in the conversion equipment 12550.
Removing these items from the conversion equipment 12550 lowers the
amount of power needed by the conversion equipment, which is
important when power is a very scarce resource as it is in a
cross-connect box.
[0096] The conversion equipment 12550 further comprises
multiplexer/inverse multiplexer (mux/imux) 12555 that multiplexes
the different digital information flows from DSL data channels and
from digitized POTS interfaces that terminate the native 0-4 KHz
POTS baseband channels. These information flows are then inverse
multiplexed across multiple loaded F1 feeder loops back to the
CO-side product or equipment 12560. Each F1 feeder loop (12657,
12757, and 12857) is connected between pairs of loaded loop
transceivers (LLTs) (12650/12660, 12750/12760, and 12850/12860,
respectively). In FIG. 12, F1 feeder loop 12657 arbitrarily has
three load coils 12651, 12653, and 12655. Also, F1 feeder loop
12757 arbitrarily has three load coils 12751, 12753, and 12755. In
addition, F1 feeder loop 12857 arbitrarily has three load coils
12851, 12853, and 12855. The CO-side product 12560 comprises loaded
loop transceivers (LLTs) 12660, 12760, and 12860, which are
connected to multiplexer/inverse multiplexer 12565. Or
alternatively, fewer LLTs, each transmitting and receiving symbols
on multiple loaded loops to achieve the advantages of diversity,
reduced complexity and reduced latency. For example, LLT 12660
could transmit/receive symbols sequentially on F1 feeder loops
12657, 12757 and 12857 to LLT 12650, eliminating the requirement
for LLTs 12760, 12860, 12750 and 12850. Inverse multiplexer 12565
may be connected to a high-speed backhaul 12570 that carries both
the digitized POTS voice and the DSL data in a packetized,
statistical multiplexing format. ATM's QoS mechanisms can be used
to prioritize queuing of different ATM cells onto the high-speed
backhaul 12570, which in a non-limiting case may just be a single
high-speed channel offering ATM's statistical multiplexing.
[0097] FIG. 13 shows a traditional approach to potentially offering
DSL service out of a cross-connect cabinet 13550. In FIG. 13,
customer premises 1, 2, 24, 48, 96, 144, 288, and 300 (with
reference numbers 13001, 13002, 13003, 13004, 13005, 13006, 13007,
and 13008) are connected to cross-connect cabinet 13550 over F2
distribution loops 13101. As shown in FIG. 13, the F2 distribution
loops 13101 to customer premises 13004 and 13005 contain load coils
13054 and 13055 respectively. As such, standard ADSL will not work
over these loops to customer premises 13004 and 13005, but ReachDSL
will work over loops with a single load coil. Using a
traditional-type of approach to providing DSL out of a
cross-connect cabinet 13550, each customer premises that includes a
native 0-4 KHz POTS interface plus DSL service must be associated
with its own F1 feeder loop 13501 back to a line card in a PSTN
switch 13560 or DLC. These F1 feeder loops 13501 for connecting the
native POTS service back to line cards may have multiple load coils
arbitrarily shown as load coils 13651, 13652, 13653, 13654, 13655,
13656, and 13657, with seven load coils normally being the maximum
found on a loop. Basically, a traditional approach might involve
trying to co-locate a DSLAM with a cross-connect cabinet 13550. A
DSLAM with multiple ADSL modems might be placed in a separate
pedestal 13570 outside the cross-connect cabinet 13550. Such a
DSLAM in a pedestal 13570 would also need CO-side POTS splitters to
isolate the frequency-division multiplexed 0-4 KHz native POTS
channel from the DSL channel. However, unlike standard DSL CO-side
POTS splitters that normally are located close to the POTS line
card in a switch or DLC, the CO-side POTS splitters would basically
be located in the middle of the POTS loop transmission line as
opposed to being near the end of the POTS loop transmission line.
Thus, the CO-side POTS splitters in the pedestal including ADSL
modems (i.e., a DSLAM) connected with cross-connect POTS splitters
13570 would desirably have different filtering characteristics than
standard CO side POTS splitters. Therefore, these CO-side POTS
splitters here are referred to as "cross-connected POTS
splitters."
[0098] Furthermore, another drawback of FIG. 13 is that the
cross-connect POTS splitters in item 13570 end up being connected
in series with the POTS service from the PSTN switch 13560 to the
customer premises. As a result, the native 0-4 KHz POTS service to
a customer premises may fail if circuitry failure (from an event
such as but not limited to loss of power) occurs in the pedestal
13570 or if the pedestal cable assembly 13580 fails. Because the
pedestal 13570 is external to the cross-connect cabinet 13550,
various environmental events, humans, or animals may cause the
pedestal cable assembly 13580 to be damaged. Such a loss of basic
POTS service that is considered a minimum acceptable lifeline may
result in legal liabilities for service providers.
[0099] Moreover, a traditional DSL methodology of trying to
co-locate a standard DSLAM with a cross-connect cabinet 13550 needs
a digital backhaul link that normally carries statistically
multiplexed DSL data traffic. In FIG. 13, remote HDSL transceiver
unit (HTU-R) 13715 is used for backhaul from the DSLAM. The HDSL
link may need one or more optional HDSL/G.SHDSL HRE repeater(s)
13725 and 13735 depending on the length of loop used for the HDSL
backhaul. Also, HDSL does not work over loaded loops. Thus, the
load coils need to be removed from the F1 feeder loops 13701 that
are used for digital backhaul from the DSLAM. The other items in
FIG. 13 generally are associated with providing power to the
equipment needed to support a traditional approach to co-locating a
DSLAM with a cross-connect cabinet 13550. HDSL/G.SHDSL Line
powering HTU-C 13745 is a central HDSL transceiver unit that also
provides line power for the HDSL backhaul equipment. One skilled in
the art will be aware that 48 volt power supplies 13755 and 13855
are commonly found in central offices and some other locations in
service provider networks such as in a DLC cabinet to power the DLC
TDM multiplexing equipment. 48 volt powering normally is not
immediately available in cross-connect cabinets 13550. Line
powering unit 13845 provides power to the DSLAM in pedestal 13570
over F1 feeder loops for power 13801. Backhaul element management
system (EMS) 13965, powering element management system (EMS) 13967,
and pedestal element management system (EMS) 13969 are used for
management and status monitoring of the various network equipment
elements.
[0100] From the description of FIG. 13, it is clear that a
traditional approach to providing DSL service out of a
cross-connect cabinet 13550 has many serious disadvantages. FIG. 14
shows the same ill-advised approach for providing DSL service out
of a cross-connect cabinet 14550 even in the absence of load coils.
In FIG. 14, customer premises 1, 2, 24, 48, 96, 144, 288, and 300
(with reference numbers 14001, 14002, 14003, 14004, 14005, 14006,
14007, and 14008) are connected to cross-connect cabinet 14550 over
F2 distribution loops 14101. As shown in FIG. 14, none of the F2
distribution loops 14101 to customer premises contain load coils so
either ReachDSL or standard ADSL will work between the customer
premises and the cross-connect cabinet 14550. Using a
traditional-type of approach to providing DSL out of a
cross-connect cabinet 14550, each customer premises that includes a
native 0-4 KHz POTS interface plus DSL service must be associated
with its own F1 feeder loop 14501 back to a line card in a PSTN
switch 14560 or DLC. Basically, a traditional approach might
involve trying to co-locate a DSLAM with a cross-connect cabinet
14550. A DSLAM with multiple ADSL modems might be placed in a
separate pedestal 14570 outside the cross-connect cabinet 14550.
Such a DSLAM in a pedestal 14570 would also need CO-side POTS
splitters to isolate the frequency division multiplexed 0-4 KHz
native POTS channel from the DSL channel. However, unlike standard
DSL CO-side POTS splitters that normally are located close to the
POTS line card in a switch or DLC, the CO-side POTS splitters would
basically be located in the middle of the POTS loop transmission
line as opposed to being near the end of the POTS loop transmission
line. Thus, the CO side POTS splitters in the pedestal including
ADSL modems (i.e., a DSLAM) connected with cross-connect POTS
splitters 14570 would desirably have different filtering
characteristics than standard CO-side POTS splitters. Therefore,
these CO-side POTS splitters here are referred to as
"cross-connected POTS splitters."
[0101] Furthermore, another drawback of FIG. 14 is that the
cross-connect POTS splitters in item 14570 end up being connected
in series with the POTS service from the PSTN switch 14560 to the
customer premises. As a result, the native 0-4 KHz POTS service to
a customer premises may fail if circuitry failure (from an event
such as but not limited to loss of power) occurs in the pedestal
14570 or if the pedestal cable assembly 14580 fails. Because the
pedestal 14570 is external to the cross-connect cabinet 14550,
various environmental events, humans, or animals may cause the
pedestal cable assembly 14580 to be damaged. Such a loss of basic
POTS service that is considered a minimum acceptable lifeline may
result in legal liabilities for service providers.
[0102] Moreover, a traditional DSL methodology of trying to
co-locate a standard DSLAM with a cross-connect cabinet 14550 needs
a digital backhaul link that normally carries statistically
multiplexed DSL data traffic. In FIG. 14, remote HDSL transceiver
unit (HTU-R) 14715 is used for backhaul from the DSLAM. The HDSL
link may need one or more optional HDSL/G.SHDSL HRE repeater(s)
14725 and 14735 depending on the length of loop used for the HDSL
backhaul. Also, HDSL does not work over loaded loops. Thus, the
load coils need to be removed from the F1 feeder loops 14701 that
are used for digital backhaul from the DSLAM. The other items in
FIG. 14 generally are associated with providing power to the
equipment needed to support a traditional approach to co-locating a
DSLAM with a cross-connect cabinet 14550. HDSL/G.SHDSL Line
powering HTU-C 14745 is a central HDSL transceiver unit that also
provides line power for the HDSL backhaul equipment. One skilled in
the art will be aware that 48 volt power supplies 14755 and 14855
are commonly found in central offices and some other locations in
service provider networks such as in a DLC cabinet to power the DLC
TDM multiplexing equipment. 48 volt powering normally is not
immediately available in cross-connect cabinets 14550. Line
powering unit 14845 provides power to the DSLAM in pedestal 14570
over F1 feeder loops for power 14801. Backhaul element management
system (EMS) 14965, powering element management system (EMS) 14967,
and pedestal element management system (EMS) 14969 are used for
management and status monitoring of the various network equipment
elements.
[0103] FIG. 15 shows a much better approach to providing DSL
service from a cross-connect cabinet 15550 using the preferred
embodiments of the present invention. In FIG. 15, customer premises
1, 2, 24, 48, 96, 144, 288, and 300 (with reference numbers 15001,
15002, 15003, 15004, 15005, 15006, 15007, and 15008) are connected
to cross-connect cabinet 15550 over F2 distribution loops 15101. As
shown in FIG. 15, the F2 distribution loops 15101 to customer
premises 15004 and 15005 contain load coils 15054 and 15055
respectively. As such, standard ADSL will not work over these loops
to customer premises 15004 and 15005, but ReachDSL will work over
loops with a single load coil. Using an approach that utilizes the
preferred embodiments of the present invention to provide DSL out
of a cross-connect cabinet 15550, each customer premises that
includes a native 0-4 KHz POTS interface plus DSL service no longer
requires its own, unshared segment of a POTS loop from the
cross-connect cabinet 15550 back to a line card in a PSTN switch
15560 or a DLC. These F1 feeder loops 15501 that were used for
connecting the native POTS service back to line cards may have
multiple load coils arbitrarily shown as load coils 15651, 15652,
15653, 15654, 15655, 15656, and 15657, with seven load coils
normally being the maximum found on a loop. However, these F1
feeder loops that had been used just for POTS service are freed up
when a customer adds DSL capability on his F2 distribution local
loop over which he will still be provided with a native POTS
interface through the conversion equipment. The freed up F1 feeder
loops can be used for other purposes such as, but not limited to,
providing additional loops for the inverse multiplexed backhaul
between the conversion equipment and the central office-side
equipment. Furthermore, the preferred embodiments of the present
invention do not require that the customer's native POTS service is
provided through digitized voice capability in the conversion
equipment. One or more customers may be wired such that the POTS
service is still delivered over a POTS loop directly from a CO or
DLC line card instead of through a line card associated with the
conversion equipment.
[0104] In the situation of FIG. 15, co-locating the conversion
equipment (comprising CO-side ADSL and/or ReachDSL modems, POTS
subscriber line cards (SLICs), multiplexing/inverse multiplexing
equipment, and loaded loop transceivers) allows DSL service to be
offered over F2 distribution loops with loaded or unloaded F1
feeder loops as backhaul for both DSL data and digitized POTS that
is used to support the native 0-4 KHz POTS interface to a customer
premises. Such conversion equipment including ADSL (or ReachDSL)
modems can be placed in an external pedestal 15570 that is
connected to the customer premises by being bridged onto the
subscriber loops through pedestal cable assembly 15580. Unlike
FIGS. 13 and 14, the bridged connections to pedestal 15570 in FIG.
15 do not present the same POTS service reliability risks that
occur with the series wiring of a POTS splitter through an external
cable in FIGS. 13 and 14. Thus, the configuration of FIG. 15 is
more immune to various environmental events, humans, and animals
that may cause the pedestal cable assembly 15580 to be broken. If
the conversion equipment loses power, then digitized POTS services
may go down. However, for any customers whose native POTS service
is provided from a line card in a CO or DLC that still has power,
the bridged connection of conversion equipment allows the customer
loop to still provide POTS service even though DSL service over
that customer loop will fail when the conversion equipment loses
power.
[0105] Additionally, unlike FIGS. 13 and 14 HTU-R 15715 and
repeaters 15725 are not required for backhauling the DSL service.
Instead, the loaded and/or unloaded F1 feeder loops 15701 carry
both the digitized voice and the DSL data. In addition, performance
generally will be better over a single unloaded loop than a singled
loaded loop. However, the inverse multiplexing in conversion
equipment allows utilization of multiple loaded and/or unloaded
loops for communicating the DSL data and digitized POTS voice
through backhaul unit 15845 that also provides line powering to the
conversion equipment in pedestal 15570 by using 48 volt power
supply 15855. Unlike FIGS. 13 and 14, the element management system
(EMS) for the pedestal, backhaul, and powering 15969 is integrated
together simplifying network management.
[0106] Using the concepts of inverse multiplexing, repeaterless
backhaul can be provided on loaded or unloaded loops. Generally,
extending the transmission line distance between two communication
devices lowers the potential channel capacity of the transmission
line, other things being equal. Repeaters are one solution to this
problem by keeping the data rate high by basically lowering the
distance over which digital signals have to propagate before a
clean copy of the information can be regenerated at a repeater. In
general, backhauling requires among other things meeting some
minimum data rate requirements for serving the acceptable
contention ratios of the backhaul link. The preferred embodiments
of the present invention include inverse multiplexing that can
effectively support a large bit rate capacity between the
conversion equipment and the CO-side equipment by utilizing a large
enough number of potentially low capacity loops. As the distance
that the data has to be backhauled increases, the bit rate
capacities of the loops generally decrease. However, this decrease
in the bit rate capacities of the loops can be compensated for by
using inverse multiplexing to gather together enough loops to meet
the data rate requirements for backhauling customer data given
various service level and contention criteria. Thus, the inverse
multiplexing of the preferred embodiments of the present invention
also helps to resolve the repeater problem for backhauling the DSL
data (and potentially the digitized POTS).
[0107] FIG. 16 shows more detail of the wiring problem from FIGS.
13 and 14 that results in the potential POTS reliability issue. In
FIG. 16, customer premises DSL equipment 16011 and POTS equipment
16013 via microfilter 16015 are connected to F2 distribution loop
16101. The F2 distribution loop 16101 connects the customer
premises with cross-connect box 16550. Cross-connect DSL equipment
16570 generally would be a DSLAM with multiple DSL modems that is
co-located with the cross-connect box 16550. In addition, to
provide standard filtering that keeps the high-frequency DSL
signals away from the POTS line card a POTS splitter is used
between the line card and the portion of the loop carrying DSL
signals. Because the filtering of the splitter should be optimized
for connection along the middle of the POTS transmission line
instead of near the end, the POTS splitter is different than a
normal POTS splitter, and FIG. 16 shows the POTS splitter as
cross-connect splitter 16580. Unfortunately cross-connect splitter
16580 is in series with the POTS interface and generally would be
physically connected to the cross-connect box 16550 through an
external cable that is subject to failure. Even integrating the
cross-connect splitter and the cross-connect DSL equipment 16590
into the same piece of equipment does not solve the problem as the
series wiring of the cross-connect splitter 16580 results in a
cable through which POTS service passes. The cross-connect splitter
16580 further connects over an F1 feeder loop to a POTS line card
in a switch or DLC 16560 that interfaces to the PSTN 16950. The
cross-connect DSL equipment 16570 such as a DSLAM has some type of
digital backhaul 16701 connection to a data network 16960.
[0108] FIG. 17 shows a potential wiring configuration for offering
DSL service out of a cross-connect box 17550. In FIG. 17, a
customer premises comprises customer premises DSL equipment 17011,
a microfilter 17015, and POTS equipment 17013 that allow the
customer to use distribution loop 17101 for both POTS and DSL. As
one skilled in the art will be aware, 710 connectors are commonly
used for making connections to cross-connect boxes 17550. 710
connectors 17123 and 17125 might normally be used for connecting
customer premises loops to cross-connect box 17550 even without the
addition of DSL service provided through cross-connect box 17550.
Normally, a cross-connect box has out terminals 17553 or junction
points and in terminals or junction points 17555 for the incoming
and outgoing wire pairs relative to the CO. As shown in FIG. 17,
710 connectors 17623 and 17627 might normally be used for
connecting customer premises loops to cross-connect box 17550 even
without the addition of DSL service provided through cross-connect
box 17550. 710 connector 17625 is added in between 710 connectors
17623 and 17627 to support DSL service from cross-connect box
17550. DSLAM equipment 17570 is connected to out terminals 17553 to
support the DSL modulation over the distribution loop 17101, and
also is connected to digital backhaul equipment 17715 for generally
statistically multiplexing the data from multiple DSL modems to
backhaul it to a data network 17790. DSLAM equipment 17570 needs
some form of powering and might be line powered through a
connection to 710 connector 17625. Digital backhaul also needs some
form of power and might use a connection to 710 connector 17625 for
line powering the backhaul loops. 710 connector 17627 includes
connections to POTS F1 local loop feeders 17501, some wires from
unused local POTS loops 17701 that are used for digital backhaul,
and some wires from unused local POTS loops 17801. The various POTS
local loops are connected back to POTS line cards 17560, digital
backhaul equipment 17745, and power equipment 17845. The POTS line
cards 17560 provide PSTN connectivity, while the digital backhaul
equipment provides data network 17790 connectivity.
[0109] The wiring of FIG. 17 is allowed without the use of a
splitter by utilizing the concepts of U.S. Pat. No. 6,111,936 to
Gordon Bremer, entitled "Method and Apparatus for Automatically
Detecting and Measuring Distortion in a DSL system", filed on Jan.
28, 1999, and issued on Aug. 29, 2000, which is incorporated by
reference in its entirety herein. In general, U.S. Pat. No.
6,111,936 teaches how distortion caused by a DSL signal can be
determined with the DSL signal. Then, the DSL signal and/or the
power spectral density (PSD) can be altered to suitably reduce or
eliminate the distortion effects, which normally would be reduced
by a POTS splitter. As a result of such adjustments to the DSL
signal, the POTS band noise is suitably reduced to resolve
potential adverse POTS quality issues. In addition, the DSL
performance can be improved as well as the undesirable distortion
is reduced. Thus, the wiring of FIG. 17 could be called
"splitterless" DSL, and it allows the DSL equipment to be connected
to subscriber loops without needing a POTS splitter. This
splitterless DSL configuration that utilizes the concepts of U.S.
Pat. No. 6,111,936 may be used with conversion equipment utilized
to provide indirect DSL service over loaded and/or unloaded loops.
In addition, the splitterless DSL configuration also will work with
normal DSLAMs.
[0110] Thus, the preferred embodiments of the present invention
provide advances over the existing technology for loaded loop
communication and allow deployment of DSL technology at higher
communication rates without the need for reengineering subscriber
loops to remove load coils. Such a solution allows
telecommunications service providers to offer higher grade service
to additional customers without absorbing the costs to rewire the
multitude of loaded subscriber loops in the networks of service
providers. Furthermore, the preferred embodiments of the present
invention will work while still allowing the loaded subscriber loop
to support both DSL data communication and POTS service. Various
priority arbitration mechanisms can be used to implement policies
for deciding when the 0-4 KHz baseband frequency is utilized for
DSL service or POTS service. In addition, various techniques can be
used to make DSL service appear to be always-on even though DSL
service may be halted in some situations of long duration use of
the 0-4 KHz baseband frequency for POTS service.
[0111] It should be emphasized that the above-described preferred
embodiments of the present invention, particularly, any "preferred"
preferred embodiments, are merely possible examples of
implementations, merely set forth for a clear understanding of the
principles of the invention. Many variations and modifications may
be made to the above described embodiment(s) of the invention
without departing substantially from the spirit and principles of
the invention. All such modifications and variations are intended
to be included herein within the scope of this disclosure and the
present invention and are to be protected by the following
claims.
* * * * *