WO2001035611A1 - Method and apparatus for cooperative diagnosis of impairments and mitigation of disturbers in communication systems - Google Patents

Abstract

A method that sends upstream a collection of data samples measured from a Digital Subscriber Line (401).

Description

Method and Apparatus for Cooperative Diagnosis of Impairments and Mitigation of Disturbers in Communication

Systems

CLAIM OF PRIORITY

This application claims the benefit of the filing date of the following

Provisional U.S. Patent Applications:

"SPECTRAL MANAGEMENT AND OPTIMIZATION THROUGH ACCURATE

IDENTIFICATION OF CROSS-TALK CHANNELS AND UNCERTAINTY",

application number 60/164,986, filed November 11, 1999;

"SPECTRAL MANAGEMENT AND OPTIMIZ ATION THROUGH ACCURATE

IDENTIFICATION OF CROSS-TALK CHANNELS AND UNCERTAINTY",

application number 60/181,125, filed on February 8, 2000;

"SPECTRAL MANAGEMENT AND OPTIMIZATION THROUGH ACCURATE

IDENTIFICATION OF CROSS-TALK CHANNELS AND UNCERTAINTY",

application number 60/183,675, filed on February 18, 2000;

"USE OF UNCERTAINTY IN PHYSICAL LAYER SIGNAL PROCESSING IN

COMMUNICATIONS", application number 60/165,399, filed November 11,

1999;

"SHARED COMPUTATIONAL RESOURCES FOR IMPROVED

PERFORMANCE OF A TRANSCEIVER" application number 60/215,159, filed

on June 30, 2000; and

"SYSTEM LEVEL SUPPORT FOR TRANSCEIVER PERFORMANCE", application number 60/215,680, filed on June 30, 2000. FIELD OF THE INVENTION

The field of invention relates to communications generally; and more

specifically, to improving the performance of a network by sharing resources or

information between a network perspective and a line perspective.

BACKGROUND

Overview

In the communications arena one of the biggest challenges is to overcome

crosstalk, noise, and other disturbances that interfere with signals. Whether the

signals are transmitted over wires, cable, fiber optics, wireless, or other types of

communication the signals suffer from some level of interference.

Interference in the signal may lead to certain limitations of the

communication system. For example in wireless systems, such as cellular

phones, interference may shorten the distance at which the signal can be reliably

received and the clarity of the signal. As another example, in wire systems, such

as digital subscriber lines (DSL), interference may shorten the distance at which

the signal can be reliably received, i.e., limit loop reach. Interference may also decrease the bit rate of the data being transferred. Providers of

telecommunications services recognize the need to monitor the quality of service

provided to users of their networks and to identify the causes of problems reported by their customers. This task, however, is complicated significantly by several factors. Some of these factors include: the large number of network users, the

large amount of data collected from the deployed lines, and the presence of

competing providers in the same physical line plant. The coexistence of ILECs

(Incumbent Local Exchange Carriers) and CLECs (Competitive Local Exchange

Carriers) in the same cable binders, brought about by the federally mandated

deregulation of local telecommunications markets, implies that services

deployed by one carrier may be disturbing the users of another carrier, who has

no information about the source of this disturbance.

It is thus highly desirable to sort through the collected data and determine

whether a specific line is being disturbed by external interference, such as AM

radio stations, or by internal interference, such as another DSL service, and

whether that offending service belongs to the same carrier or not. Unfortunately,

with today's deployed monitoring technology, carriers are extremely limited in

their ability to perform such diagnoses with adequate accuracy and reliability.

The following discussion outlines in detail many of the problems of digital

subscriber line (DSL) technology and potential solutions thereto. However, the

discussion merely uses DSL as one example of the many communication systems (e.g., wireline, wireless, optical, cable, etc.) in which the present invention may be used. Thus the present invention should not be limited to merely DSL

communication systems.

Overview with Respect to DSL Digital Subscriber Line (DSL) networks provide high speed networking

service while preserving the investment made in traditional telephone lines.

Figure 1 shows an exemplary topology of a DSL network. In the exemplary DSL

network topology 100 of Figure 1, various customer premise equipment (CPE) modems 105, 106, 107 are communicatively coupled to a central office switching center 101 via ordinary telephone lines (e.g., lines 120 through 122).

Customer premise equipment 105, 106, 107 is equipment located at the

customer's location (e.g., a customer's home or office). In the exemplary network

topology 100 of Figure 1, the customer premise equipment 105, 106, 107

possesses at least one transceiver (e.g., transceiver 108 in CPE 105) that is

responsible for: 1) controlling at the CPE the reception of information sent from

the service provider; and 2) controlling at the CPE the transmission of

information sent to the service provider.

Information that flows in the network 100 toward the customer (e.g.,

toward the direction of a CPE as seen in Figure 1) has a "downstream" direction while information that flows in the network 100 away from the customer (e.g.,

away from a CPE as seen in Figure 1) has an "upstream" direction. Thus it may

be said that a transceiver within a CPE is responsible for controlling at the CPE

the transmission of upstream information and the reception of downstream

information.

Various DSL service schemes exist. For example, at a high level, DSL services are characterized according to the bandwidth allocated for a customer's upstream and downstream traffic. Services that reserve approximately equal

amounts of bandwidth for a customer's upstream and downstream traffic are

referred to as "symmetric DSL" while services that reserve approximately

unequal amounts of bandwidth for a customer's upstream and downstream

traffic are referred to as "asymmetric DSL".

Symmetric DSL (SDSL), High bit rate DSL (HDSL, HDSL-2) and ISDN

DSL (IDSL) are versions of symmetric DSL. Asymmetric DSL (ADSL), Rate

Adaptive DSL (RADSL), Very high bit rate DSL (VDSL), and G.Lite are versions

of asymmetric DSL. Any of these DSL services (as well as other potential future

DSL services that are not listed above) may be referred to as "DSL".

Note that the central office 101 includes a plurality of DSL Access

Multiplexers 102, 103, 104 (DSLAMs). A DSLAM operates as a distributor of

DSL services. That is, for example, DSLAM 102 forwards /collects

downstream /upstream information sent from /to higher layers of a service

provider's network to/from transceivers 108, 109; 110.. The service provider's

DSL network is controlled by a Network Management Agent (NMA) 118.

An NMA 118 is one or more software routines that monitor the operation

of a network (e.g., by collecting various performance monitoring statistics sent from the DSLAMs 102, 103, 104) and controls various aspects of a network (e.g., by enabling or disabling service on a particular line). The NMA 118 shown in Figure 1 monitors and controls the DSL network 100 by communicating with the

DSLAMs through the Element Management Systems 116, 117 (EMSs). The NMA 118, as an example, may be executed as part of a network's Network

Management System (NMS). An EMS effectively distributes to the DSLAMs

control information sent from the NMA and forwards to the NMA 118 network

performance or network status indicia sent from the DSLAMs. More details on a

DSL system are provided below.

Figure 2 shows an exemplary depiction of a receiver 201 within a DSL

transceiver 208. That is, for example, transceiver 208 of Figure 2 may be viewed

as corresponding to transceiver 108 of Figure 1 and line 220 of Figure 2 may be

viewed as corresponding to line 120 of Figure 1. Recalling that the transceiver

208 is responsible for controlling both the transmission of upstream traffic and

the reception of downstream traffic, note that receiver 201 assists the

performance of the latter of these two functions.

The receiver 201 includes an equalizer 202 and a symbol detection unit 203

(which may also be referred to as a symbol detector 203). The equalizer 202 adjusts the transfer function of the receive channel such that the frequency

components of the received waveform rx(t) 221 that are associated with the

signal (i.e., the frequency components of the received waveform rx(t) 221 that are

associated with the downstream information sent from the service provider to

the transceiver 208) are enhanced with respect to the frequency components of the waveform rx(t) 221 that are not associated with the signal (i.e., the frequency

components of the waveform's "noise"). For example, the signal components alone may be amplified, the noise components alone may be suppressed or a

combination of both.

The symbol detection unit 203 converts the features of the equalized

waveform 222 into digital Is and Os according to the modulation scheme

employed by the particular type of DSL service being implemented. As a result

of the equalizer's activity, the signal-to-noise ratio (SNR) in the receive channel is

enhanced and the performance of the symbol detection unit 203 (i.e., its ability to

correctly reproduce the digital information sent by the service provider) is

improved.

Referring back to Figure 1, note that the ordinary telephone lines that

couple the DSLAMs and the CPEs are tightly packed together in a binder such as binder 114 and binder 115. Because ordinary telephone lines were originally

designed for low speed voice /telephony communications, they are typically

packed in a binder without shielding. That is, a line is not protected from

receiving electromagnetic interference associated with the waveforms that

appear on another line; nor are the waveforms on a line prevented from

radiating so as to interfere with the waveforms that appear on another line.

The interference described above, commonly referred to as cross-talk, is viewed as noise that may corrupt the operation of the symbol detection unit 203 discussed above with respect to Figure 2. Cross-talk typically increases as the frequencies of the waveforms on an ordinary telephone line increase. When the ordinary telephone lines were originally installed to carry voice

traffic, cross-talk was insubstantial because of the lower frequencies used to

transmit voice traffic. However, as DSL is designed to provide higher speed

services (as compared to traditional telephony service) over these ordinary

telephone lines, cross-talk from DSL waveforms is much more severe. The more

severe cross-talk frequently hampers the successful deployment of a DSL service.

SUMMARY OF INVENTION

A method that sends upstream a collection of data samples measured from a network line.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is illustrated by way of example, and not limitation,

in the Figures of the accompanying drawings in which:

Figure 1 shows an exemplary DSL network topology;

Figure 2 shows an exemplary DSL receiver within a DSL transceiver-

Figure 3a shows an exemplary line perspective;

Figure 3b shows an exemplary network perspective;

Figure 4 shows an improved DSL receiver having a cross-talk compensation unit;

Figure 5 shows an embodiment of a DMT- ADSL DSL receiver that conforms to the improved DSL receiver approach of Figure 4;

Figure 6 shows a methodology for developing a line perspective;

Figure 7 shows a depiction of event notification flows that may be used to

develop a network perspective;

Figure 8 shows a depiction of information being shared between a line level perspective and a network level perspective;

Figure 9 shows another depiction of information being shared between a line level perspective and a network level perspective; Figure 10 illustrates an exemplary communication system 1005 that may benefit

from the present invention; and

Figure 11 illustrates a DSL system.

DETAILED DESCRIPTION

1.0 Overview

In the following description, for purposes of explanation, numerous

specific details are set forth in order to provide a thorough understanding of the present invention. It will be evident, however, to one skilled in the art that the

present invention may be practiced without these specific details. In some

instances, well-known structures and devices are shown in block diagram form,

rather than in detail, in order to avoid obscuring the present invention. These

embodiments are described in sufficient detail to enable those skilled in the art to

practice the invention, and it is to be understood that other embodiments may be utilized and that logical, mechanical, electrical and other changes may be made

without departing from the scope of the present invention.

Some portions of the detailed descriptions that follow are presented in

terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations

are the means used by those skilled in the data processing arts to most effectively

convey the substance of their work to others skilled in the art. An algorithm is

here, and generally, conceived to be a self-consistent sequence of acts leading to a

desired result. The acts are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values,

elements, symbols, characters, terms, numbers, or the like.

It should be borne in mind, however, that all of these and similar terms

are to be associated with the appropriate physical quantities and are merely

convenient labels applied to these quantities. Unless specifically stated

otherwise as apparent from the following discussion, it is appreciated that

throughout the description, discussions utilizing terms such as "processing" or

"computing" or "calculating" or "determining" or "displaying" or the like, refer to

the action and processes of a computer system, or similar electronic computing

device, that manipulates and transforms data represented as physical (electronic)

quantities within the computer system's registers and memories into other data

similarly represented as physical quantities within the computer system

memories or registers or other such information storage, transmission or display

devices. The present invention can be implemented by an apparatus for

performing the operations herein. This apparatus may be specially constructed for the required purposes, or it may comprise a general purpose computer, selectively activated or reconfigured by a computer program stored in the

computer. Such a computer program may be stored in a computer readable

storage medium, such as, but not limited to, any type of disk including floppy disks, optical disks, CD-ROMs, and magnetic-optical disks, read-only memories

(ROMs), random access memories (RAMs), EPROMs, EEPROMs, magnetic or optical cards, or any type of media suitable for storing electronic instructions,

and each coupled to a computer system bus.

The algorithms and displays presented herein are not inherently related to

any particular computer or other apparatus. Various general purpose systems

may be used with programs in accordance with the teachings herein, or it may

prove convenient to construct more specialized apparatus to perform the

required method. For example, any of the methods according to the present

invention can be implemented in hard-wired circuitry, by programming a

general purpose processor or by any combination of hardware and software. One

of skill in the art will immediately appreciate that the invention can be practiced

with computer system configurations other than those described below,

including hand-held devices, multiprocessor systems, microprocessor-based or

programmable consumer electronics, network PCs, minicomputers, mainframe

computers, and the like. The invention can also be practiced in distributed computing environments where tasks are performed by remote processing

devices that are linked through a communications network. The required

structure for a variety of these systems will appear from the description below.

The methods of the invention may be implemented using computer

software. If written in a programming language conforming to a recognized

standard, sequences of instructions designed to implement the methods can be compiled for execution on a variety of hardware platforms and for interface to a variety of operating systems. In addition, the present invention is not described with reference to any particular programming language. It will be appreciated

that a variety of programming languages may be used to implement the

teachings of the invention as described herein. Furthermore, it is common in the

art to speak of software, in one form or another (e.g., program, procedure,

application...), as taking an action or causing a result. Such expressions are

merely a shorthand way of saying that execution of the software by a computer

causes the processor of the computer to perform an action or produce a result.

2.0 Overview of General Communication Network

The present invention is applicable to a variety of communication

systems, for example: wireline, wireless, cable, and optical. Figure 10 illustrates

an exemplary communication system 1005 that may benefit from the present

invention. The backbone network 1020 is generally accessed by a user through a

multitude of access multiplexers 1011 such as: base stations, DSLAMs (DSL

Access Mulitp lexers), or switchboards. The access multiplexers 1011

communicate management data with a Network Access Management System

(NAMS) 1010. The NAMS 1010 includes several management agents 1015 which are responsible for monitoring traffic patterns, transmission lines status, etc. Further, the access multiplexers 1011 communicate with the network users. The user equipment 1040 exchanges user information, such as user data and

management data, with the access multiplexer 1011 in a downstream and

upstream fashion. The upstream data transmission is initiated at the user equipment 1040 such that the user data is transmitted from the user equipment 1040 to the access multiplexer 1011. Conversely, the downstream data is

transmitted from the access multiplexer 1011 to the user equipment 1040. User

equipment 1040 may consist of various types of receivers that contain modems

such as: cable modems, DSL modems, and wireless modems.

The invention described herein provides a method and system for

managing the upstream and downstream data in a communication system. As

such, the present invention provides management agents that may be

implemented in the NAMS 1010, the access multiplexers 1011, and /or the user

equipment 1040. One example of such a management agent is a system software

module 1070 that may be embedded in the NAMS 1010. Another management

agent that manages the data in the communication system 1005 is a transceiver

software module 1060 that may be embedded in the access multiplexer 1011 and/or the user equipment 1040. Further details of the operation of modules 1070 and 1060 are described below.

For illustration purposes and in order not to obscure the present

invention, an example of a communication system that may implement the

present invention is a DSL communication system. As such, the following

discussion, including Figure 11, is useful to provide a general overview of the

present invention and how the invention interacts with the architecture of the

DSL system.

3.0 Overview of DSL Example The present invention may be implemented in software modules or

hardware that DSL equipment manufacturers may then embed in their

hardware. Thus, although Figure 11 illustrates the present invention as software,

the present invention should not be limited thereto. It should also be noted that

this patent application may only describe a portion or portions of the entire

inventive system and that other portions are described in co-pending patent applications filed on even date herewith.

Figure 11 illustrates an exemplary embodiment of the present invention as

implemented in a DSL system. The DSL system consists of a network of

components starting from the Network Management System (NMS) 1110 all the

way down to the Customer Premise Equipment (CPE) 1150. The following is a brief description of how these components are interconnected.

The Network Management System (NMS) 1110 is a very high level

component that monitors and controls various aspects of the DSL system

through an Element Management System (EMS) 1120. The NMS 1110 may be

connected to several Central Offices (CO) 1111 through any number of EMSs

1120. The EMS 1120 effectively distributes the control information from the NMS

1110 to the DSL Access Multiplexers (DSLAMs) 1133 and forwards to the NMS 1110 network performance or network status indicia from the DSLAMs 1133.

DSLAMs 1133 reside in a Central Office (CO) 1111, usually of a

telecommunications company. Alternatively, DSLAMs 1133 may reside in remote enclosures called Digital Loop Carriers (DLC). The CO 1111 may have tens or hundreds of DSLAMs 1133 and control modules (CM) 1132. A DSLAM

1133 operates as a distributor of DSL service and includes line cards 1135 and

1136 that contain CO modems. The CO modems are connected to at least one

line 1145, but more frequently it contains several line cards 1135 and 1136 that

are connected to several lines 1145. Usually the lines 1145 are traditional phone lines that consist of twisted wire pairs and there may be multiple lines 1145 in a

binder 1140 and multiple binders in a cable. The transmission cables act as

packaging and protection for the lines 1145 until the lines 1145 reach the

Customer Premise Equipment (CPE) 1150. It should be noted that a DSLAM

1135 does not necessarily have to be connected to lines 1145 in a single binder

1140 and may be connected to lines in multiple binders 1140. The lines 1145

terminate at the CPE 1150 in transceivers that include CPE modems. The CPE

1150 may be part of or connected to residential equipment, for example a

personal computer, and /or business equipment, for example a computer system

network.

As discussed in the background section, communications systems often suffer from interference and /or impairments such as crosstalk, AM radio, power

ingress noise, thermal variations, and /or other "noise" disturbers. The present

invention or portions of the present invention provide the user the capability to

analyze, diagnose and/or compensate for these interferences and/or

impairments. It also provides the ability to predict and optimize performance of the communication system in the face of impairments. As illustrated in Figure 11, the transceiver software of the present

invention 1160, depending upon how implemented, may provide the user with

the ability to analyze, diagnose, and compensate for the interference and /or impairment patterns that may affect their line.

Also as illustrated in Figure 11, the system software of the present

invention 1170, depending upon how implemented, may provide the service

provider with the ability to diagnose, analyze, and compensate for the

interference and /or impairment patterns that may affect the service they are

providing on a particular line. The diagnosis and analysis of the transceiver

software also provide the ability to monitor other transmission lines that are not

connected to the DSLAMs or NMS but share the same binders.

It should be noted that the system software of the present invention 1170

may be implemented in whole or in part on the NMS 1110 and/or EMS 1120

depending upon the preference of the particular service provider. Likewise, it should be noted that the transceiver software of the present invention 1160 may

be implemented in whole or in part on the DSLAM 1133 and/or transceivers of

CPE 1150 depending upon the preference of the particular user. Thus, the

particular implementation of the present invention may vary, and depending

upon how implemented, may provide a variety of different benefits to the user

and /or service provider.

It should also be noted that the system software of the present invention 1170 and the transceiver software of the present invention 1160 may operate separately or may operate in conjunction with one another for improved

benefits. As such, the transceiver software of the present invention 1160 may

provide diagnostic assistance to the system software of the present invention

1170. Additionally, the system software of the present invention 1170 may

provide compensation assistance to the transceiver software of the present invention 1160.

Thus, given the implementation of the present invention with respect to

the DSL system example of Figure 11, one of ordinary skill in the

communications art would understand how the present invention may also be

implemented in other communications systems, for example: wireline, wireless,

cable, optical, and other communication systems. Further details of the present

invention are provided below. Additional examples of how the present

invention may be implemented in a DSL system are also provided below for

illustrative purposes.

4.0 Overview of a Line Perspective and a Network Perspective

Recall from the background that cross-talk between lines in a DSL network may hamper the successful deployment of the DSL network. Figures 3a

and 3b relate to perspectives of a DSL network that may be developed by two different network components (e.g., a transceiver 308 as seen in Figure 3a and the

NMA 318 as seen in Figure 3b) in order to understand the causes and /or effects of cross-talk within the DSL network. Figure 3a shows a perspective that may be developed at the line of a DSL

network (e.g., by a DSL transceiver 308). A line perspective is a collection of

information that characterizes the environment of a DSL line. For example, the

line perspective of line 320 in Figure 3a includes a model for each source of cross-

talk noise that disturbs signal reception on line 320. A source of cross-talk noise

(e.g., a waveform on a proximate line) may be referred to as a disturber. Cross¬

talk noise may therefore also be referred to as disturber noise. The exemplary

line perspective of Figure 3a indicates that the DSL transceiver 308 has identified

three different disturbers dl(t), d2(t), and d3(t).

Thus, for example, disturber dl(t) may correspond to a waveform on a

first line, d2(t) may correspond to a waveform on a second line, and d3(t) may

correspond to a waveform on a third line. Each disturber dl(t), d2(t), d3(t)

passes through and is processed by its corresponding co-channel hl(t), h2(t), and

h3(t). Each co-channel hl(t), h2(t), and h3(t) represents the impulse response of

the electromagnetic coupling that exists between lines that "cross-talk" with one

another.

In order to improve the signal-to-noise ratio (SNR) in the transceiver's

receive channel, as an example, the transceiver 308 may develop as part of its line

perspective of line 320: 1) the disturber signals dl(t), d2(t), d3(t); and 2) each disturber's corresponding co-channel impulse response hl(t), h2(t), and h3(t). With this line perspective of line 320, the transceiver 380 may then at least approximate and remove disturber noise on line 320. As a result, SNR will be improved. More details as to how a line perspective may be developed are

provided further below.

Figure 3b shows another perspective of a DSL network that may be

referred to as a network perspective. A network perspective is an understanding

of cross-talk (or other interference) as developed through the correlation of

information obtained from events observed on the lines within a network. As

seen in Figure 3b, note that a network perspective may be developed by a DSL

network's NMA 318. The NMA 318 "keeps track of" events such as changes in

the performance and /or configuration of each line in the DSL network that the

NMA 318 exhibits control over.

By keeping track of and correlating information from these events, an

understanding of the cross-talk between the network's lines (or other types of

interference) may be developed. For example, if the NMA 318 observes as an

event that lines 320, 321 and 323 each experienced a drop in SNR just after an

increase in the service speed on line 324 was allowed, the NMA 318 can assume

that line 324 presents disturber noise on lines 320, 321 and 322.

Furthermore, in light of the amount of SNR reduction that has occurred

on each line, the NMA 318 can approximate the degree of cross coupling between line 324 and each of lines 320, 321 and 322. This information may be

used, for example, to prevent or limit the extent of further speed increases on line 324 (so that proper operation of lines 320, 321 and 322 is ensured). More details as to how a network perspective may be developed is provided further below. The discussion that follows demonstrates that transferring line

perspective information to a network perspective and /or transferring network

perspective information to a line perspective may result in the development of

more accurate line and /or network perspectives and /or increased performance

of the DSL network as a whole. For example, referring to Figure 3b, if the

network perspective is informed that the line perspective of line 321 includes a

disturber having a frequency that corresponds to the speed of line 324, the

network perspective's confidence that line 324 presents cross-talk on line 321 is

increased.

As a second example, recall that a line perspective may develop an

understanding of the disturbers dl(t), d2(t), d3(t) and their corresponding co-

channels hl(t), h2(t), and h3(t). The computational resources available to a

transceiver 308 that develops this line perspective (e.g., a microprocessor or digital signal processor (DSP) in the CPE) may be limited so that only a less detailed perspective of the disturbers and /or their corresponding co-channels

can be developed.

By measuring the waveform rx(t) and /or its associated frequency spectra

and then forwarding this data upstream from the transceiver 308 (e.g., to the

DSLAM, EMS or NMA) more accurate models of the disturbers and /or their co-

channels can be developed. Typically, the DSLAM, EMS, and NMA possess

more powerful computational resources (e.g., a multi-processor processing core) than the CPE. By executing disturber and co-channel identification routines on these

more powerful computational resources, more precise characterizations of the

disturber and co-channel profiles may be developed. The more precise

characterizations may then be sent downstream to the transceiver 308 resulting

in an improved SNR as compared to the SNR that would have been achieved by

executing disturber and /or co-channel characterization sequences at the CPE

alone.

The following discussion develops a deeper understanding of how the

line and network perspectives of Figures 3a and 3b may be developed. A

discussion of the information that may be exchanged between these perspectives,

and the benefits derived from this exchange, follows immediately afterward.

Before continuing, however, it is important to emphasize that the techniques,

design strategies, improvements, advantages, etc. discussed below may be

applied to network technologies other tha DSL (e.g., wireless networks, fiber

optic networks, etc.). As such, a line may be viewed more generally as a

communication channel that exists between a service provider and customer

(e.g., a wireless link, a fiber optic cable, a copper cable, etc.).

5.0 Development Of A Line Perspective

A. Improved Receiver Operation

Before discussing a methodology for developing a line perspective, the design and operation of an embodiment of a DSL receiver that is able to improve SNR by maintaining a line perspective will first be discussed. Figure 4 shows an

embodiment 401 of a DSL receiver as described just above. The operation of the

equalizer 402 and the symbol detection unit 403 of Figure 4 corresponds to the

operation of the equalizer 202 and symbol detection unit 203 as described with

respect to Figure 2.

Comparing the prior art DSL receiver 201 of Figure 2 with the improved

DSL receiver 401 of Figure 4, note the insertion of a cross-talk compensation unit

490 between the equalizer 402 and the symbol detection unit 403. The cross-talk

compensation unit 490 includes a signal removal unit 404, disturber receiver 407

and a disturber removal unit 408. The cross-talk compensation unit 490 removes

disturber noise from the received waveform on line 420 prior to the symbol

detection that is performed by the symbol detection unit 403. Referring to

Figures 3a and 4, the signal removal unit 404 and the disturber receiver 407

together represent a channel that provides a representation (e.g., a time domain

representation or a frequency domain representation) of one or more disturber

signals (e.g., disturber signals dl(t), d2(t), d3(t) of Figure 3a) to the disturber removal unit 408 at a first disturber removal unit input 450.

The disturber removal unit 408 accepts the disturber signal

representation(s) and effectively processes them according to: 1) their

corresponding co-channel hl(t), h2(t), h3(t); and 2) the activity of the equalizer 402. This processing produces a representation of the disturber noise as it

appears at the output of the equalizer 403. The disturber removal unit 408 then combines (e.g., subtracts) the disturber noise representation with the equalizer

402 output to produce an equalized signal having reduced disturber noise. The

equalized signal having reduced disturber noise is then presented to the symbol

detection unit 403 so that the signal (i.e., the downstream information sent from

the service provider) may be detected.

Recall that the signal removal unit 404 and the disturber receiver 407

together represent a channel that provides a representation (e.g., a time domain

representation or a frequency domain representation) of one or more disturber

signals (e.g., disturber signals dl(t), d2(t), d3(t) of Figure 3a) to the disturber

removal unit 408 at a first disturber removal unit input 450. The signal removal

unit 404 removes those aspects of the equalizer 403 output that correspond to the

signal being sent as downstream traffic from the service provider to the receiver

401.

That is, to the extent possible, the output of the signal removal unit 404

corresponds to pure "noise". The disturber receiver 407 includes a disturber

equalizer 405 and a disturber symbol detection unit 406. The disturber equalizer

407 attempts to "undo" the activity of the equalizer 203. That is, recall from the discussion in the background that an equalizer (such as equalizer 203 of Figure 2

or equalizer 403 of Figure 4) suppresses a channel's noise and /or amplifies it's

signal.

In so doing, the equalizer 203 attempts to "whiten" the noise so that it possesses an approximately constant power spectral density over the frequency range of interest. As a result, particularly strong disturber noise frequency

components (e.g., a 20 - 392 KHz band for a symetric DSL service on a

neighboring line) will be individually and disproportionately attenuated by the

equalizer 203 (as compared to other noise frequency components). The disturber

equalizer 405 attempts to reverse this disproportionate attenuation so that the

pure spectral components of the disturber noise, as they appear on the line 420

prior to processing by the equalizer 402, are recaptured.

After the original disturber noise power profile is approximately

recaptured by the disturber equalizer 405, the disturber symbol detection unit

406 reconstructs or otherwise deduces (to the extent possible), one or more

disturber signals (e.g., disturber signals dl(t), d2(t), d3(t) shown in Figure 3a) as

they appear on their respective lines. These are then forwarded to the disturber

removal unit 408.

Figure 5 shows an exemplary embodiment of a Discrete Multi Tone -

Asymmetric Digital Subscriber Line (DMT- ADSL) receiver 501 that conforms to

the processing approach just described with respect to Figure 4. The DMT- ADSL receiver 501 of Figure 5 includes an equalizer 502 (which corresponds to the equalizer 402 of Figure 4), a DMT signal removal unit 504 (which corresponds to

the signal removal unit 404 of Figure 4), a disturber receiver 507 (which

corresponds to the disturber receiver 407 of Figure 4), a disturber removal unit 508 (which corresponds to the disturber removal unit 408 of Figure 4) and a symbol detection unit 503 (which corresponds to the symbol detection unit 403 of

Figure 4).

During a sequence referred to as "line training", the equalizer 502 searches

for the signal based upon the type and /or speed of service that is to be received. When the signal is found, the equalizer 502 adjusts an impulse response function

profile associated with a time domain equalizer (TEQ) 509. This impulse

response function, when convoluted with the received signal rx(t), produces an

efficient representation of the received signal rx(t) at the TEQ output 509.

Furthermore, the TEQ convolution may also provide (as an ancillary benefit)

some degree of noise suppression.

An efficient representation of the received signal rx(t) may be realized by

limiting the number of samples used to represent the convolution of the TEQ

impulse response with the channel g(t) impulse response. An efficient

representation of the received signal rx(t) reduces the processing load presented

to the transceiver's processing resources (e.g., a microprocessor or digital signal processor or combination of both) for subsequent processing of the received

waveform rx(t).

The fast Fourier transform (FFT) unit 517 converts the efficient

representation of the received waveform rx(t) from the time domain to the

frequency domain. The frequency domain equalizer (FEQ) 518 searches for and extracts an efficient frequency domain representation of the DMT signal originally transmitted by the service provider tx(t). The DMT signal removal unit 504 corresponds to the signal removal unit

404 of Figure 4. As such the output 510 of the DMT signal removal unit 504

corresponds to, to the extent possible, pure "noise". As seen in Figure 5, a slicer

unit 521 detects (within the frequency domain) the DMT signal.

DMT is a modulation scheme that uses a plurality of quadrature

amplitude modulated (QAM) sinusoids to transmit digital information. The

frequency of each sinusoid is centered within a frequency "bin" (e.g., a frequency

band of 4.3125 KHz reserved for its transmission. According to QAM

modulation, the phase and amplitude of a sinusoid are modulated to represent the different possible states of the digital bits being transmitted. The number of

bits that are transmitted on a line increases with the number of sinusoids that are

transmitted and /or the number of different phase and amplitude positions (i.e.,

bit states) implemented per sinusoid.

The slicer unit 521 effectively determines the phase and amplitude of the

received sinusoid (s) in order to extract the symbol information sent by the

service provider. Remodulator 522 constructs a DMT signal (as modified by the

channel g(t) and the TEQ 509) from this symbol information. The DMT signal is then converted to the time domain by the inverse Fourier transform (IFFT) unit

519. Pure noise is created by subtracting the DMT signal (as presented by the IFFT unit 519) from the equalizer output 522. Pure noise, in this case, may also be referred to as a DMT-compensated signal. It is important to point out that the receiver approach outlined in Figure 5

may be used for DSL signals other than a DMT- ADSL signal. That is, the

receiver design of Figure 5 may be tailored for any DSL service provided the

proper modulation schemes are accounted for. For example, if the received

signal is a Pulse Amplitude Modulated (PAM) signal (as is the case with SDSL

signals), the DMT removal unit 504 can be reconfigured as a PAM removal unit if

a PAM based slicer 521 is employed. Thus, even though the particular

embodiment being discussed with respect to Figure 5 is limited to a DMT- ADSL

application, those of ordinary skill will recognize that the receiver approach of

Figure 5 is actually applicable to DSL services other than DMT- ADSL.

After the DMT signal is removed by the DMT signal removal unit 504,

disturber signals are generated by the disturber receiver 507. The disturber

receiver 507 includes a disturber equalizer (DEQ) 505 (that corresponds to the

disturber equalizer 405 of Figure 4) and a disturber symbol detector 506 (that corresponds to the disturber symbol detector.406 of Figure 4). In the

embodiment of Figure 5, the disturber symbol detector employs Viterbi

processing techniques and thus may also be referred to as a Viterbi detector 506.

As discussed with respect to the disturber equalizer 405 of Figure 4, the DEQ 505 attempts to "undo" any noise suppression provided by the equalizer

502. That is, with respect to the design approach of Figure 5, the DEQ 505 attempts to undo any noise suppression provided by the TEQ 509. Noise

suppression from the TEQ may be undone by effectively inversely compensating for the adjustments made by the TEQ (during line training as

discussed above) to the TEQ impulse response function profile.

That is, recalling that the TEQ 509 provides noise suppression, if the DEQ

505 impulse response function adjustment is opposite to that of the TEQ 509, the

noise suppression provided by the TEQ 509 may be effectively eliminated. As

such, at the DEQ 505 output, the disturber noise as it appears on the line has

been re-captured.

In an embodiment, the DEQ 505 is designed using minimum -mean-

squared-error (MMSE) techniques. The result of this DEQ design is filter

coefficients that a yield a filter which effectively counteracts the noise suppression provided by the TEQ 509.

The Viterbi detector 506 of Figure 5 corresponds to the disturber symbol

detector 406 of Figure 4. As such, the Viterbi detector 506 reconstructs or

otherwise deduces (to the extent possible), one or more disturber signals (e.g.,

disturber signals dl(t), d2(t), d3(t) shown in Figure 3a) as they appear on their

respective lines.

For example, if a particular portion of the disturber noise presented by the DEQ 505 is understood to be a PAM-SDSL signal that is cross coupled to the

receiver's line 520, the Viterbi detector 506 reconstructs the PAM-SDSL signal on the cross coupled (e.g., nearby) PAM-SDSL line from that portion of the disturber noise. Similarly, if another particular portion of the disturber noise presented by

the DEQ 505 is understood to be a second PAM-SDSL signal that is cross coupled to the receiver's line 520, the Viterbi detector 506 may also reconstruct the second

PAM-SDSL signal on the second cross coupled (e.g., nearby) PAM-SDSL line

from the other portion of the disturber noise.

Note that the association of particular portions of the disturber noise

presented by the DEQ 505 with specific types of "nearby" services is an aspect of

the line level perspective (discussed with respect to Figure 3a) held by the

receiver 501. An exemplary embodiment of how this understanding /perspective is developed is provided in more detail below. The Viterbi detector 506 of Figure

5 employs Maximum Likelihood Sequence Estimation (MLSE) to reconstruct,

from its line level perspective, the particular disturber signal on the cross

coupled line.

For example, PAM signals are used to transmit two bits of information via the modulation of a pulse amplitude (e.g., a pulse amplitude of +3 may

correspond to 11, a pulse amplitude of +1 may correspond to 10, a pulse

amplitude of -1 may correspond to 01, and a pulse amplitude of -3 may

correspond to 00). The Viterbi detector 506 employs an MLSE technique to determine whether particular portions of the disturber noise presented by the

DEQ 505 (that are understood to be caused by a particular PAM disturber) correspond to a +3, +1, -1, or -3. As such, the particular sequences of +/-3 and

+/-1 deduced by the Viterbi detector 506 correspond to a disturber signal on a

cross coupled line. The deduced disturber signal is then presented to the disturber removal

unit 508 at the disturber receiver output 550. Other disturber signals that the

Viterbi detector 506 is designed to detect are also presented at the disturber

receiver output 550. Note that the number of disturber signals that the Viterbi

detector 506 is designed to detect is an aspect of the line level perspective held by

the receiver 501.

For example, in an embodiment, the number of disturber signals detected

by the Viterbi detector are limited (e.g., to 2 or 3 typically) by processing

limitations available to the receiver 501. Thus, in such an embodiment, part of

the Viterbi design process involves prioritizing which disturbers cause the

strongest disturber noise.

After the disturber signals are presented to the disturber removal unit 508,

the disturber remodulator 515 effectively passes each disturber signal through an

estimation of its corresponding co-channel. That is, referring briefly back to

Figure 3a, recall that a line level perspective may include not only an

understanding of a disturber signal (e.g., disturber signals dl(t), d2(t), and d3(t))

but also an understanding of the channel (referred to as a co-channel hl(t), h2(t),

h3(t)) that the cross-talk passes through in reaching the affected line 320. A

discussion of how a co-channel may be estimated for each disturber signal is

presented in more detail below.

The disturber remodulator 515 convolves the disturber signals dl(t), d2(t), d3(t) that are received from the disturber detector 507 with the impulse response function of its corresponding co-channel (e.g., hl(t) for dl(t), h2(t) for d2(t), and

h3(t) for d3(t)). As a result, a representation of the disturber noise that is

produced on the line 520 from these disturbers is created. This representation is

then convolved with an impulse response h_TEQ(t) that is representative of the

equalization provided by the TEQ 509.

As a result, a representation of the disturber noise as it appears at the

output of the TEQ 509 is created. This disturber noise representation (which

corresponds to the disturber remodulator 515 output) is then converted from the

time domain to the frequency domain by an FFT unit 516. The output of the FFT

516 is then multiplied by the FEQ 528 (which may be identical to the FEQ 518).

This signal is, to the extent possible, identical to the disturber noise as it is

presented at the output of the FEQ 518. By subtracting the output of the FEQ 528

from the FEQ 518 output, the disturbers will be approximately canceled (i.e.,

removed from the FEQ 518 output). As such, the SNR is enhanced before the

symbols on the line 520 are detected by the symbol detection unit. It is important to point out that the receiver 501 approach outlined above should not be

construed as limited to the particular frequency domain processing / time domain processing strategy that is displayed in Figure 5.

B. Development of Line Level Perspective During Line Training

Recall from the discussions above that a line level perspective may be

developed that includes: 1) an understanding of the disturber signals that exist on one or more cross coupled lines (e.g., as represented by disturber signals dl(t), d2(t), and d3(t) in Figure 3a) and their corresponding co-channels (e.g., as

represented by impulse responses hl(t), h2(t), and h3(t) in Figure 3a).

Figure 6 shows a methodology that may be used to develop the line

perspective discussed just above. The development of a line perspective (and /or

any disturber noise compensation that results) may be referred to as mitigation of disturbers. The development of a line perspective may occur during line

training. Line training is a period of time prior to the actual use of the line to

transmit a customer's information (referred to as "showtime"). During line

training the CPE transceiver responsible for controlling the

transmission /reception of upstream /downstream traffic "learns" about the

operating environment of the line.

For example as seen in Figure 6, before showtime occurs, the equalizers

(e.g., the TEQ 509 and FEQ 518 of Figure 5) undergo a training sequence in which

the proper adjustments for suppressing the line's noise and /or amplifying the line's signal are established. After the equalizer adjustments are established, the

disturber signals and their corresponding co-channels may be identified and /or otherwise characterized 610.

Then, based on the understanding of the disturber signals and their

corresponding co-channels: 1) the disturber receiver (e.g., disturber receiver 407

of Figure 4) is tailored 620 to detect the particular disturber signals chosen for compensation. This process is completed before showtime begins. Referring to Figure 6, a disturber signal may be identified or otherwise

characterized through its type of service 601. Said another way, with foresight of

the types of services that may cause disturber noise (e.g., TI or PAM-SDSL on a

DMT-ADSL line), certain frequency ranges may be "focused upon" to see if

disturber noise exists.

That is, for example, it is known that a TI signal has a fundamental

frequency of approximately 1.5 MHz. By searching across a frequency range

centered at 1.5 MHz, the existence (or lack thereof) of disturber noise resulting

from a cross-coupled TI line may be confirmed and its exact frequency may be

determined. Such a frequency range may be referred to as a "service specific" frequency range.

If disturber noise power (e.g., above a critical threshold to warrant further

analysis) is detected for a particular service type, the corresponding frequency

range may be further analyzed 602 to see how many disturber signals (i.e., how

many cross coupled lines) exist for this type of service. For example, by

"refocusing" in the service specific frequency range with a finer resolution

bandwidth, each discovered "peak" may be assumed to be caused by a different

line (owing to differences in crystal oscillator frequencies used to form the disturber signals). Note that identification of the frequency at which a particular

peak occurs corresponds to a further refinement of the line level perspective. That is, not only has the service type for a source of disturbance been identified but also its particular frequency has been identified. Once the number of disturbers of a particular service type is determined, a

model of the spectral content of an ideal disturber signal for each discovered

disturber is generated. This ideal disturber signal model may be compared

against what is actually observed on the line (i.e., the disturber signal's

corresponding disturber noise) to generate 603 an estimation of the disturber signal's co-channel. That is, the co-channel is responsible for (and may be

characterized by) the "distortion" that occurs to the disturber signal as it is

converted from a disturber signal to disturber noise.

The process described just above is iterated until each of the potential (i.e.,

foreseen) types of service that can cause disturber noise on the line are analyzed

(e.g., when all the service specific frequency ranges have been searched over).

Note that the concept of service type may be extended to include any cause of

disturber noise. As such the method described above should not be construed as

limited only to networking services that exist on an ordinary telephone line. For example, AM radio station carrier frequencies may be searched for any resulting

disturbance noise.

Note that once all of the disturber sources and their co-channels have been

identified, a complete line level perspective has been developed. That is, the

service type and frequency particular to each disturber signal has been identified. Furthermore, the profile particular to each corresponding co-channel has also been identified. The disturbers are then ranked according to disturber noise power. That

is, the highest powered observed disturber noise is ranked first, the second

highest powered observed disturber noise is ranked second, etc. As such, a

corresponding ranking of disturber signals results. The disturber signals chosen

for compensation are taken from the ranking (coextensive with the processing

constraints that apply).

The line level perspective is then built 620 into the design of the

transceiver. First, because the amount of disturber noise that will be removed is

understood, the transceiver can estimate its expected improved SNR and

determine 606 an appropriate line speed (or data rate) for the line. Second, the

disturber equalizer (e.g., disturber equalizers 405, 505 of Figures 4 and 5) is

configured 607 to "undo" the equalization of the equalizer (e.g., equalizer 402,

502 of Figure 4 and 5) based upon the parameters that setup the equalizer.

Third, the disturber symbol detector is configured 608 to detect the chosen

disturber signals according to the particular type of service that have been

identified for each disturber. The co-channel for each disturber signal and the

impulse response of the equalizer is also made available to the disturber removal

unit. This process is completed before showtime.

For examples of the methodologies and apparati discussed just above, see co-pending patent applications entitled "Method and Apparatus for

Characterization of Disturbers in Communication Systems", "Method and Apparatus for Mitigation of Disturbers in Communication Systems", "Design & Architecture of an Impairment Diagnosis System for Use in Communication

Systems", "Method and Apparatus for Impairment Diagnosis in Communication

Systems", "Method and Apparatus for Prediction & Optimization in Impaired

Communication Systems" all assigned to the present assignee and filed on an

even date herewith.

6.0 Development of Network Perspective

Figure 7 shows a depiction of event notification flows that may be used to

develop a network perspective. Development of a network perspective (and/or any network improvement that results) may be referred to as diagnosis of

impairments. Recall from the discussion of Figure 3b that a network perspective

is an understanding of cross-talk (or other interference) as developed through the

correlation of events observed on the lines within a network. The cross-talk

understanding may be embodied in the form of a "chart" that identifies: 1) which

lines are cross coupled with one another; and 2) for each cross coupling that is identified, how strong the particular cross coupling is.

A networking perspective may be developed by designing intelligence at the line level (e.g., within a CPE transceiver) that sends notification of an event upstream to higher layers of the network (such as the NMA 718). For example, an event may be designed to correspond to: 1) an observed change in SNR on a

line (or a change in bit error rate or other measurement of signal quality); 2) a change in transmitter signal power on a line as demanded or otherwise

permitted by the NMA; and /or 3) a change in transmitted bit rate speed for a line as demanded or otherwise permitted by the NMA. The direction of the

event notification flow is indicated by the arrow heads seen in Figure 7.

In an embodiment, event notifications are sent to and collected by the

NMA 718. The NMA "keeps track of" these events and attempts to correlate

them with other network events that the NMA is aware of. For example, if the

NMA 718 collects event notifications from lines 720, 721 and 722 that each has

experienced a drop in SNR just after an increase in the service speed on line 724

was allowed, the NMA 718 can assume that line 724 is the source of disturber

noise on lines 720, 721 and 722.

Furthermore, in light of the amount of SNR reduction that has occurred

on each line, the NMA 718 can approximate the strength of the cross coupling

that exists between line 724 and each of lines 720, 721 and 722. This information may be used, for example, to prevent or limit the extent of further bit rate increases on line 724 (so that proper operation of lines 720, 721 and 722 is

ensured).

When event notifications are sent upstream from a CPE transceiver to a

line card, the events are collected at the line card that is responsible for communicating with the line that experiences an event. Thus, for example, if line

card 706 communicates over lines 720, 721 and 722, line card 706 collects the

events from these lines.

An event notification may be time stamped by a transceiver before it is sent upstream or may be time stamped by the line card that receives the event notification. By comparing the timestamps of the collected event notifications,

the line card is able to assume that some events are correlated while other events

are not correlated. Specifically, those events that occurred at approximately the

same time may be deemed as related to one another (e.g., by a cause/effect

relationship). For example, if a drop in SNR occurs at approximately the same

time on lines 720, 721, 722, the line card may assume that each of these SNR

changes had the same cause.

Each line card 706, 707, 708 reports its event notifications (and any

correlation it has discovered) to a DSLAM control unit 704. The DSLAM control

unit prioritizes and condenses the event information before sending them further

upstream to the NMA 718. For example, insignificant events (such as a small change in SNR) may be ignored by the DSLAM control unit 704.

As another example, with respect to DSLAM 702, consider an instance

where each line card 706, 707, 708 reports an event that was correlated to each of

the line card's lines. If the DSLAM control unit 704 further determines that each

of these reports are correlated (e.g., the timestamp of the event reported by each

line card 706, 707, 708 are approximately the same), the DSLAM control unit 704 may conclude that a "significant" event has occurred that has affected every line coupled to the DSLAM 702.

The reporting of this event to the NMA may take priority over (i.e., be sent prior to) other events that have already been reported by the line cards but do not correlate to as many lines. Queuing of events at the DSLAM control unit 704 may be employed if the flow of events to be reported exceeds the bandwidth

of the communication link between the DSLAM 702 and the NMA 718. The

DSLAM control unit 704 may also be configured to condense the event

information (e.g., by coupling multiple events in a single message to the NMA)

to enhance the efficiency of upstream event notification flow.

The NMA 718 collects the network events sent by the DSLAMS it has

control over. The NMA 718 performs a higher level correlation by correlating

events reported by different DSLAMs. For any change in SNR caused by an

adjustment in line power or line speed within the DSL network, the NMA 718

will be able to get a full report of the lines affected and build an understanding of

the cross-talk that exists in the network. Furthermore, specific verification tests may be executed to see if a particular line change is allowable.

For example, if an existing customer desires to increase the speed of his or

her service, the service provider can send a higher speed test signal over the line.

Depending on the SNR changes to other lines that are caused by the increase in speed, the service provider may permit or deny the increased service.

Furthermore, the service provider may continuously run tests during a

network's "quiet time" (e.g., in the early morning when the customer population

is mostly asleep). By continually running tests (e.g., adjustments in speed

and /or power to one or more lines) and continually collecting the events that

follow, the NMA 718 can build upon and improve its understanding of the crosstalk that exists between the lines on its network. For examples of improving an understanding through continued

monitoring and analysis of the lines see patent applications entitled "Design and

Architecture of an Impairment Diagnosis System for Use in Communication

Systems" and "Method and Apparatus for Impairment Diagnosis in

Communication Systems" assigned to the present assignee and filed on an even

data herewith.

That is, the aforementioned "chart" (that identifies: 1) which lines are

cross coupled with one another; and 2) for each cross coupling that is identified,

the strength of the particular cross coupling) is continuously refined and

improved as to its accuracy. Note that so far the network perspective has been limited to "in domain" lines. In domain lines are lines that the NMA 718 has

control over (in terms of being able to adjust their speed or power) and can

receive event notifications from.

The NMA 718 may also be able to build an understanding of "out of

domain" disturbers (i.e., disturbers that the NMA 718 does not have control over

and does not receive event reports from). For example, if a local AM radio station reduces its transmitted power every day after sunset, those in-domain

lines that are cross coupled with the AM radio station will report an increase in

SNR every day after sunset. The NMA 718 can therefore add to the "chart" the existence of an AM radio station that affects the lines that indicate cross coupling. Various other processes may also be used to identify at least the presence of disturbers originating from lines that are controlled by other service providers. Other types of diagnosis and analysis reports may be generated,

depending upon the application. For instance, in a DSL application, a report

may include the type of activity of all diagnosed out-of-domain and in-domain

disturbers disturbers and victims that are estimated to be in a given binder.

Since twisted pair lines in a binder often terminate in a small geographic area of users, e.g., within several hundred feet, such a report may also provide

information regarding services deployed by other carriers in that small

geographical area.

7.0 Exchange of Information Between A Line Level Perspective and A

Networking Level Perspective

The discussion that follows demonstrates that transferring line

perspective information to a network perspective and /or transferring network

perspective information to a line perspective may result in the development of a

more accurate perspective and /or increased performance by a DSL network as a whole. For example, referring to Figure 8, if a description 801 of the disturber

sources observed at a line are sent upstream (e.g., to an NMA 818 that also

collects event notifications from lines in the network) a number of improvements

may be realized.

First, the NMA 718 may develop a more accurate "chart" of lines that are cross coupled. That is, recall that the disturber information gathered during a

line perspective development phase (as discussed with respect to Figure 6) includes: 1) description of the service that the disturber signal corresponds to; 2)

the actual frequency of the disturber signal; and 3) an estimate of the co-channel

between the line and the cross coupled line carrying the disturber signal.

By sending such a description of one or more disturbers upstream to the

NMA 818, the NMA 818 can more readily and with more confidence develop its

chart. For example, if the NMA 818 believes (as a result of the event reporting

described above with respect to Figure 7) that a particular line (e.g., line 724 in

Figure 7) causes disturber noise on particular lines (e.g., lines 720, 721, and 722),

this belief may be "confirmed" if the victim lines (e.g., lines 720, 721, and 722)

each report identical disturber information that match the configuration of the

disturber line (e.g., line 724).

That is, if line 724 is configured to deliver a 784kbps PAM-SDSL service and lines 720, 721 and 722 each send a disturber profile corresponding to a

784kbps PAM-SDSL service with the same actual frequency, the NMA 818 may establish with a very high degree of confidence that line 724 is cross coupled with

lines 720, 721, and 722.

Furthermore, recall that the event reporting scheme discussed above with

respect to Figure 7 allowed the NMA 818 to develop an understanding of the

strength of the cross coupling that exists between lines. The accuracy of the

coupling strength understanding may be enhanced if the disturber information 801 sent upstream to the NMA 818 also includes a description of a disturber's co- channel. That is, the co-channel provides a thorough description of the cross coupling's dependency on disturber signal frequency. By the event reporting

process alone, frequency dependent information is mostly gained by actually

changing the operational speed of a line (e.g., via a test signal).

With this collection of information, the network may be better optimized

by the service provider. That is, the service provider can predict with improved

accuracy the effect that an increase in bit rate or an increase in transmitted power

will have on the service provider's other lines. As such, the service provider is

more able to correctly allow or deny such increases (if requested by a customer)

based upon the actual understood cross coupling that exists among the service

provider's lines.

Note that disturber information 801 sent upstream to the NMA may also

describe "out-of domain" disturbers. Thus, whereas the event reporting scheme

may be limited to realizing only the existence of an "out of domain" disturber,

the sending of out of domain disturber information to the NMA 818 allows the

NMA 818 to gain a deeper understanding of the out of domain disturber.

Specifically, the type of service, the service speed and the co-channel of the out of domain disturber may be understood.

With this information, the NMA may be able to confidently configure networking service arrangements that will not be affected by the out of domain disturber. For example, due to regions of overlapping and non overlapping frequency usage by various DSL services, it is understood that a DMT-ADSL

service will interfere (i.e., introduce disruptive disturber noise) with an SDSL service but not with a CAP-ADSL service. As a result, if the NMA 818 gains an

understanding from the disturber information 801 (sent from the line

perspective) that an out of domain disturber corresponds to a DMT-ADSL

service, the NMA 818 may be configured to allow CAP- ADSL service on the

victimized lines but not an SDSL service. Similar judgments may be exercised

based upon the understood speed of an out of domain disturber and the speed of a desired service or service upgrade that may be offered by the NMA 818.

Furthermore, depending on the presence or absence of non volatile

memory within a CPE, additional DSL network enhancement may be gained if

the CPE uses the NMA 818 (or other upstream equipment such as a DSLAM or

EMS) as an effective non volatile storage unit. That is, a CPE without non

volatile storage will lose its line level understanding if its power is turned off.

When the CPE is turned on again, the entire line level understanding will have to be re-developed.

However, if the CPE can use the NMA as its non-volatile storage, it may

be able to maintain (and even improve) its line perspective. In particular, after

the CPE initially builds its line perspective information, it can forward it to the

NMA 818 through an upstream management data channel. In the case of DMT- ADSL, this corresponds to an Embedded Operations Channel (EOC).

The NMA 818 can store this line perspective information for later re¬

transmission, in addition to using it to improve its own network perspective. When that particular CPE is turned on after being turned off, it can request that

its line perspective information be sent back to it from the NMA.

By receiving this information, the CPE does not have to devote

computational resources to rebuilding its line perspective. Instead, it can use

those resources to further refine its line perspective information (e.g., by

collecting more data and focusing even more closely on the specific frequencies

of the detected disturbers).

Also, if the NMA 818 notices a change in the disturber profile (i.e., if a

significant loss in SNR is reported by various lines), the NMA 818 may request

any line to "re-develop" its line level understanding. When the subsequent disturber information 801 gathered from the new line understanding is

forwarded to the NMA 818, the NMA 818 can search for the cause of the change

(e.g., such as a newly introduced out of domain disturber).

Figure 9 relates to another demonstration that transferring line

perspective information to a network perspective and /or transferring network

perspective information to a line perspective may result in the development of a

more accurate perspective and/or increased performance by a DSL network as a

whole. In the demonstration of Figure 9, note that data samples 901 taken from a line are sent upstream to more sophisticated equipment in the network (such as the equipment used to implement the NMA 918, an EMS or a DSLAM).

Referring back to Figure 6, recall that the transceiver observed the spectral content on its line in order to develop its understanding of the disturbers that affect the line. In the demonstration of Figure 9, observed spectral content of the

line (i.e., data samples 901) is forwarded upstream. Upstream equipment may

process these data samples with a methodology the same as, similar to or

different from the line perspective development methodology 610 outlined in

Figure 6.

That is, because upstream equipment tends to have more powerful computational resources, a more detailed and accurate analysis of the spectral

content on the line may be performed (as compared to the analysis performed at

the CPE). For example: 1) a wider service specific frequency range may be used;

2) a narrower resolution bandwidth that searches for a disturber peak may be

employed; 3) the entire frequency spectrum (rather than just service specific

frequency ranges) may be scanned; 4) a more robust algorithm for detecting the

particular service type may be used (e.g., an algorithm that scans for the

presence of higher frequency harmonics); 5) a more precise co-channel estimation

may be developed, etc. The results of any these analyses will take the form of a more accurate disturber description 902 which is then forwarded back to the

line's transceiver.

As an alternate cooperative enhancement, note that the NMA's development of a more detailed and accurate analysis of the spectral content of

the line may be used to "pinpoint" to the CPE transceiver precisely where

important disturbers are to be found. That is, for example, the disturber information 902 directed to the CPE may be used by the CPE to execute its own (i.e., "local") transceiver training and design routines (e.g., as discussed with

respect to Figure 6). Because the NMA has informed the CPE transceiver "where

to look", the CPE transceiver can immediately focus upon one or more

disturbers, rather than scan a wide frequency range. More efficient use of

training time results (e.g., via improved disturber and co-channel models and /or

reduced time spent during training the training period).

Note also that these more advanced disturber descriptions may be made

accessible to the NMA, thus at least all of the advantages and improvements

discussed above with respect to Figure 8 may be realized. Furthermore, recall

from Figure 6 that after the line understanding is developed 610 it is built into

the design 620 of the transceiver. The more powerful computational resources of the NMA may also be used to calculate more precise design parameters for the

disturber equalizer and the disturber symbol detector. The more precise design

parameters 904 may be sent back to the CPE transceiver 908 so that they may be

integrated into its design.

In particular, the NMA may perform an improved ranking of disturbers, not according to their power, but according to the severity of the impairment

that they cause on the victimized line of the CPE. This ranking typically requires

much more sophisticated and computationally expensive processing. As a

result, a more suitable set of disturbers may be selected for compensation, with a corresponding increase in CPE compensation performance. Additionally, the NMA may employ much more sophisticated algorithms

for computing the optimal filter coefficients for the DEQ. Such algorithms may

include least-squares or min-max methods that require more memory than is

available on a typical CPE processing resource. This optimized DEQ design

results in further improvements in CPE computational performance.

Claims

CLAIMSWhat is claimed is:

1. A method, comprising:

sending upstream a collection of data samples measured from a DSL line.