JP4891779B2 - Adaptive margin control and adaptive bandwidth control - Google Patents

Description

TECHNICAL FIELD The present invention relates generally to methods, systems, and apparatus for managing digital communication systems. More specifically, the present invention is not limited to these, but in a communication system such as a DSL system, maximum transmission power spectral density, maximum total transmission power, transmission band priority, minimum and maximum receiver margin, frequency It relates to adaptive control of various transmission parameters including dependent bit loading and power control and / or bit loading restrictions.

Background of the Invention Digital subscriber line (DSL) technology can provide a potentially large bandwidth for digital communications over existing telephone subscriber lines (referred to as line and / or copper plant). Telephone subscriber lines can achieve this bandwidth despite their original design for voice band analog communications only. Specifically, Asymmetric Digital Subscriber Line (ADSL) uses a discrete multi-tone (DMT) line code that assigns a number of bits to each tone (or sub-carrier) to address subscriber line characteristics. Which can be adjusted at each end of the subscriber line by channel conditions determined during training and initialization of a modem (typically a transceiver that functions as both transmitter and receiver). Can be adjusted against. Adaptive allocation uses a secure, relatively slow reverse channel and uses a process called “bit-swapping” to inform the transmitter of the allocation change, over time, on a channel or line Can continue during raw data transmission.

  Impulse noise, other noise and other sources of error can significantly affect the accuracy of data transmitted by ADSL and other communication systems. Various methods have been developed to reduce, avoid and / or repair data damage due to such errors during transmission. These error reduction / avoidance / repair techniques have a performance cost relative to the communication system in which the technique is used. As is well known in the art, insufficient power transmission levels cause errors because the transmitted power is not sufficient to overcome noise and other interference in a given channel. These errors lead to data loss and / or the need for multiple retransmissions of data. To prevent such errors, the system uses extra transmit power that results in a margin above the known or predicted signal-to-noise ratio (SNR) that ensures adherence to acceptable error rates.

  In general, a pair of DSL modems determines the power, margin and other operational characteristics of the pair during initialization, training, channel analysis and exchange phases (sometimes referred to as SHOWTIME). The process begins with a base power spectral density (PSD) value or mask. This can be “flat” or a constant value (ie frequency independent), or it can be a variable mask, where the PSD value is a frequency specific or frequency dependent value. It is. Various DSLs have an initial PSD value (sometimes referred to as “NOMPSD”), and in general, the upper limit of NOMPSD is defined by standards applicable to any country. Using this initial PSD value, the modem predicts line attenuation and line length (and possibly other parameters and / or values).

  Using the line attenuation and line length prediction, one or both modems can define a power reduction or power reduction (PCB) value that reduces the initial PSD value. As shown below, different DSL standards set and adjust power (eg, PSD and PCB) according to different rules when the standard is actually being followed and complied with.

Once the adjusted PSD value (REFPSD = PSD-PCB) may be used, the modem calculates the bit loading (b _i ), gain (g _i ), and margin during the channel analysis phase. The gain g _i is an adjustment to the transmit power level of individual bit-loaded tones in the DMT scheme and provides a relatively uniform margin for the transmission of data on the line. The gain can be adjusted during SHOWTIME to reflect line state changes and the like, but may be limited depending on the amount of adjustment and the way in which such gain adjustment can be performed.

  Depending on the manufacturer of the equipment and the DSL standard used, compliance with the rules and guidelines of each standard and / or other appropriate operational restrictions may be strictly enforced by several parties (this is usually the DSL service rate during operation). Ranging from a very modest or timid setting) to a casual disregard for basic operational guidelines and rules. In many cases, to avoid problems that may be caused by lack of power and / or margin, whether intentional or unintentional non-compliance, excess power and A margin is used.

  However, excessive transmission levels cause other problems. For example, the use of one or more lines of excessive transmit power can cause strong crosstalk and interference in adjacent lines. Crosstalk is unwanted interference and / or signal noise that travels electromagnetically between lines that share the same binder or adjacent binders. Also, the use of transmission power exceeding the required level means that the communication system is operating at a very high cost, damaging all users. The following is a summary of the current standards and implementations for some aspects of the DSL service to which embodiments of the invention disclosed below can be applied, with each specific standard and the more general Includes implications for distinguishing between initial setup, training, channel analysis and exchange procedures.

ADSL-G. 992.1 standard (also referred to herein as "ADSL1 standard" or "ADSL1")
(1) The standard has a MAXSNRM maximum margin (or equivalent) limit setting that can be specified by the operator, but the ability to comply with and enforce this limit will vary depending on the modem manufacturer and interpretation of the standard, As a result, it is often virtually ignored. In general, an “operator” is a telecommunications company or other service provider that operates a network and provides the service itself. Internet service providers are generally not considered operators because they typically subcontract services to other parties.

  (2) ATU-R (downstream receiver) and ATU-C (upstream receiver) are limited to a maximum gain reduction requirement of 14.5 dB, which is insufficient to implement the MAXSNRM objective It may be. In addition, some ATU-R modems ignore MAXSNRM and actually verify that the margin does not exceed MAXSNRM. There has never been an interoperability test that states non-compliance with 992.1.

  (3) When the downstream ATU-C transmitter indicates that the upstream signal received by the initial training is large and the line is short, the G. Reduce to 12 dB according to the algorithm of the 992.1 Annex (the most popular are Annexes A, B and C). Annex A, B, and C algorithms are not sensitive to downstream channel output noise, so they are very bold when reducing power, and in most cases never reduce power sufficiently. For example, the widely used Annex A algorithm applies only to lines that are 3000 feet or less in length, and thus in many situations where longer lines may require power reduction as well. Cannot adapt.

(4) The upper limit or “mask” of the initial flat PSD can be programmed according to the MAXNOMPSD parameter, which is −40 to −52 dBm / Hz, in 2 dB steps. The MAXNOMPSD value is set by the operator, and NOMPSD is the maximum value that can be assumed at the start of transceiver training. The modem manufacturer can configure the modem to use a lower NOMPSD value, where NOMPSD <MAXNOMPSD. The MIB / Telecom operator could only set the MAXNOMPSD in the initial system, but could not enforce the NOMPSD value itself. In ADSL1, NOMPSD is communicated by the transmitter to the receiver (for later attenuation calculations) during training. In this case as well, there is no MIB parameter that directly specifies NOMPSD in ADSL1. Generally, MAXNOMPSD is the NOMPSD used in the very early transceiver training phase of ADSL1, but the NOMPSD level (which is limited to be below MAXNOMPSD) is determined by the modem manufacturer, not the operator. . Thus, by setting MAXNOMPSD, the operator can reduce the upper limit level of NOMPSD to -2n _PCB dBm / Hz as defined when n _PCB = 0-6 (ie, when NOMPSD is -40) , PSD power can be reduced by 0, 2, 4, 6, 8, 10 or 12 dB).

  (5) Some manufacturers' ATU-R receivers completely ignore the MAXSNRM and 14.5 dB despite such power reduction being mandated by the operator and by the standards. No power reduction is required.

(6) Gain swapping dynamic Sakuchu, after training in SHOWTIME, + / - 2.5 dB limit of to allow only gain adjustment (and will be recommended, not required in Annex A) may be. That is, gain swapping is limited to a total of +/− 2.5 dB with ADSL1 in some modems. If the gain is not reduced enough for some reason during training (eg, spurious noise was present), no further reduction will occur unless the modem is retrained. Retraining record (retraining count) is an indication that how many times retraining has been performed within a given period, and that the MAXSNRM level can be set very low if the retraining count is too high Can be stored by the DSL system. Gain swaps are applied continuously in ADSL1 (as they establish each other), but any SHOWTIME continuous gain reduction of gain swaps is limited to a maximum of +/− 2.5 for training. There is a case. However, some vendors require a series of gain reductions that lead to a -14.5 dB full ADSL1 allowable gain reduction in SHOWTIME. If a change below −2.5 dB is desired, the receiver in such a full-range gain swap system adjusts the internal signal processing (unlike the other 68 raw data symbols, during a SHOWTIME due to a gain swap. It does not have to be sufficiently powerful to prevent intersymbol interference from certain synchronization symbols (which is not reduced during operation). Some inadequate receivers require power reduction, so that if a gain reduction is implemented that is lower than a gain reduction of -2.5 dB (and the operator places where those inadequate receivers are located) The receiver itself cannot be adjusted internally). And such inadequate ATU-R considers the line to be degraded, rather it should be addressed if the problem is due to incomplete implementation of ATU-R requiring gain reduction. DSL connection is dropped. Because of this problem, the service provider, the DSLAM provider, the dynamic Sakuchu, + / - power reduction request exceeding the range of 2.5dB always good even Forces to ignore (and service providers, Because you don't know where insufficient receivers are located, forcing it throughout the network will limit all good receivers) . In addition, this limits power reduction if the service provider chooses this option that does not require changing insufficient receivers already in place in the network.

ADSL2-G. 992.3 standard (also referred to herein as "ADSL2 standard" or "ADSL2")
(1) The above standard has a MAXSNRM setting, but this function remains in the DSL receiver for implementation.

(2) The ATU-R (as a downstream receiver) and ATU-C (as an upstream receiver) are limited to a 14.5 dB maximum power reduction requirement for gain setting, and the gain setting is , Absolutely set in gain swap and not executed for the last gain swap. Since the range here is −14.5 to [+ 2.5 + EXTGI], the range is limited to a maximum power reduction of 14.5 dB (EXTGI ≧ 0 and generally equal to 0, EXTGI Is something that tells the receiver during initial training, which can be adapted during later gain swapping). Greater EXTGI up to the limit of 18dB is a modem having a power that is reduced for some reason, to increase its power to the dynamic Sakuchu, corresponding to the new more noise that may occur during operation Make it possible.

  (3) A PCB in ADSL2 improves the ability to comply with MAXSNRM because the receiver allows the receiver to reduce power by an additional 0, 1,..., 40 dB (during training only). The ADSL2 standard can realize that the maximum PCB required by either the transmitter or receiver should be implemented. MAXNOMPSD is the only parameter managed by the operator in ADSL2 that applies to the entire band, but a wider range of this parameter is applied in ADSL2 than in ADSL1.

  (4) The initial flat PSD mask can be programmed according to the MAXNOMPSD parameter, which is 0.1 dB steps at −40 (and −37 in some extended annexes with ADSL2 known as READSL) ) To -60 dBm / Hz.

  (5) Some manufacturers' ATU-R receivers may ignore the PCB, and the receivers are unfortunately not tested even in the new WT Forum test procedure WT-85. Yes, but almost all tests go through the receiver, and MAXSNRM compliance is not confirmed in the test). In the ADSL2 standard itself, band priority (which is a frequency-dependent obligation of PSDMASK) is not possible.

(6) Gain swapping in S HOWTIME is no longer limited to +/− 2.5 dB and all symbols (note that there is a sync symbol every 69th) have the same level. However, only reduced gain swapping to -14.5 dB is possible with respect to the training level (as in ADSL1, not the last level). Gain increases up to 2.5 + EXTGI are particularly useful when the modem starts up with very low power and noise occurs. If EXTGI is large, the modem can recover without retraining. EXTGI is limited to 18.0 dB in ADSL2.

ADSL2 + -G. 992.5 standard (also referred to herein as "ADSL2 + standard" or "ADSL2 +")
(1) Same as ADSL2, except for the introduction of PSDMASK implemented by the tssi parameter. tssi is an additional parameter similar to the gain in gain swapping, except that it can be fixed externally.

VDSL1, VDSL2, HDSL and SHDSL
The current version of the proposed VDSL1 standard or 993.1 has a limited definition of MIB-managed (or operator-controlled) procedures for power reduction (DSLAM or line terminal (LT) modem manufacturers have built-in full PSDMASK However, access to it via MIB is currently unclear in G.993.1). G. A list of 993.1 standard maintenance capabilities appears in TR-057 of the DSL Forum document, but the MIB management chapter of TR-057 is currently blank. Thus, VDSL does not have a standardized mechanism for setting MAXNOMPSD externally, but can be set programmatically in a nominally imposed specification limit (G.993.1 compliant modem 2 Two mask levels and corresponding transmit power levels are downstream and upstream, so power reduction is for them, some of which are not yet specified) It has an internal mechanism for reducing power in 0.25 dB steps (called manual power control) between 0 and 40 dB, and between 0 and 12 dB if downstream. VDSL also specifies MAXSNRM (but again it is not clear who will specify this). Thus, VDSL has many of the same capabilities as ADSL1 and ADSL2 / 2 +. Those capabilities that allow many similar performances as ADSL1, ADSL2, and ADSL2 + may be standardized for operator interfaces in the MIB in future documents. However, because VDSL1 does not have a rich set of ADSL2 and ADSL2 +, or apparently ADSL1's published diagnostic methods, the ability to accurately diagnose problems can be more difficult. Again, the next generation TR057 or G.I. 997. x may correspond to those defects in the current VDSL MIB interface.

  It is clear that VDSL2 is still in a very early stage, but will have essentially the same MIB characteristics as ADSL2 +. HDSL is not obvious in any configuration with some power reduction characteristics. HDSL (currently upgraded to SHDSL, G991.2) has a target SNR (or TSNRM) and a published SNRM but does not have a MAXSNRM. The bandwidth is symmetrical and can be one of several data rates (basically 384, 768, 1.5, 3, ...) with the same bi-directional modulation as some standardized shaping. It is fixed against. A flat PCB of 0,..., 31 dB can be imposed. Since there is no FEC to protect against impulses, it is undesirable to use PCBs in excess. Also, since SHDSL tends to operate at the maximum possible rate over short lines, its margin will typically approach 6 dB TSNRM.

  DSM reports that are still in the draft stage currently have all MIB capabilities of ADSL2 + in both directions and apply to all DMT transmission methods, ADSL1 to VDSL2, and beyond. FEC can also be specified.

  As will be appreciated by those skilled in the art, in many DSL systems, including ADSL1 and ADSL2 systems, the operating characteristics and rules are generally relative to the static mode of operation to accommodate the worst case scenario in the system. Is set. That is, users do not always realize the full benefits of a DSL system due to inadequate standards, equipment limitations, and lack of generally accepted processing procedures and specifications. For example, power margin limits are rarely observed, can be inconsistent with various standards or interpretations of those standards, or can be inconsistent between various standards or interpretations of those standards. is there. Such disregarding the limits imposed by service providers and / or standards establishes problems for users, including excessive crosstalk. Similarly, impulse noise can be a significant problem in some DSL systems. To address impulse noise, current systems use default settings supplied by the manufacturer for a number of operating parameters (such as margins). Applicable standards are intended to allow service providers to set their parameters, but it is not currently known whether the standards are properly implemented by various vendor DSL modems or devices.

  Even if the majority of users in the binder (ie their modems) are compliant, only one user is the source of service degradation or other damage to other users' DSL services. May be found. For this reason, the standard provides guidance, but in the current system, slight non-compliance can pose a serious problem.

  Static operation (eg, if the DSL service uses the default settings set by the manufacturer in the DSL modem), the DSL service changes the line and environmental conditions in the subscriber line, and / or Means that it cannot be adjusted and / or adapted for reducing the profits gained with a DSL system and not being able to realize effective potential for one or more users in such a system . As will be appreciated by those skilled in the art, widely changing standards, devices, implementation rules (or lack thereof) and implementation are consistent services and detailed standards for the operation of these various DSL systems. It means that service quality is difficult. The modems and other devices may not actually follow or comply with the appropriate standard, and more importantly, the user's adjacent line should not be used with standard compliant devices and implementations. Many users receive services that are insufficient or less than optimal.

  Systems, devices, methods and techniques that allow users to dynamically adjust and adapt transmit power margin, power spectral density, etc. to changing DSL environments and operating conditions are It will be a significant advance in the field. Also, monitoring and evaluation of power, margins, etc. used in DSL environments and operations by independent entities can assist with user activity and equipment guidance and (in some cases) management, as well as DSL There will be significant progress in the field of operations.

Summary of the Invention Controlling margins in a DSL modem pair is based on collected operational data. The operational data is analyzed and at least one modem in the modem pair is instructed to use margin-related parameter values calculated so that the modem pair meets a margin target, such as a margin limit imposed by the DSL standard or the like. Embodiments of the present invention can be used with ADSL1, ADSL2, ADSL2 +, VDSL and other types of DSL systems and devices.

  Controllers such as DSM centers, “high performance” modems and / or computer systems can be used to collect and analyze the operational data. The controller and / or other components can be a computer-implemented device or combination of devices. In some embodiments, the controller is remote from the modem. In another case, the controller can be placed with one or both of the modems as a device directly connected to one modem, thereby forming a “high performance” modem.

The margin related parameter value may be a PSD related value such as a MAXNOMPSD or MAXNOMATP parameter used in various ADSL systems. In some embodiments, the margin related parameter value may be a shaped spectral mask used for transmission and / or an upper limit or limit on bit loading for the frequency used for transmission between the modems. can do. In some cases, a priority band can be provided to instruct the modem to prefer and / or avoid certain frequencies.

  The operational data may include historical data related to previous performance and previous margin compliance of the modem pair. The operating data may also include one or more modem operating parameters whose values are the same as or different from the margin related parameters adjusted by the controller. Historical data can be stored in a database. The operational data may further include data collected from the DSL system, in which case the modem pair functions based on, for example, one or more MIBs or other data sources. The operational data can be sent to the controller by communication means internal or external to the DSL system itself. Some other types of operational data that can be evaluated include data on crosstalk between the modem pair and the adjacent DSL line, history of margin previously used by the modem pair (if the retraining count level is high) Data relating to retraining count, transmit power level previously used by the modem pair, data rate previously used by the modem pair, and / or previous error effects of the modem pair. including.

  In another embodiment of the invention, the distribution of margins based on operational data can be estimated as a function of data rate. Using the estimated margin distribution, a distribution of performance parameters is also calculated, including the possibility of line outages and the possibility of one or more parameters exceeding a minimum level. Data rates and / or other performance-related parameters can be set based on the estimated performance of the system using various margin settings and levels.

  The method according to the invention can be performed by a controller, a DSM center, a “high performance” modem, a computer system or the like. Also disclosed are computer program products for performing those methods.

  Further details and advantages of the present invention are set forth in the following detailed description and associated drawings.

  The present invention will be readily understood by the following detailed description and the accompanying drawings, and like reference numerals designate like components.

DETAILED DESCRIPTION OF THE INVENTION The following detailed description of the invention refers to one or more embodiments of the invention, but is not limited to such embodiments. Rather, the detailed description is intended to be illustrative only. Those skilled in the art will readily appreciate that the detailed description set forth herein with respect to the drawings is set forth for purposes of explanation, since the scope of the invention extends beyond those exemplary embodiments. Will.

  It should be noted that the details described herein are for illustrative purposes and that the present invention is broader in scope than any one embodiment. As such, the present invention should be construed as broadly as possible and allowed.

  In general, embodiments of the present invention include controllers (eg, computer systems, “high performance” modems, dynamic spectrum managers, spectrum management centers (Spectrum Management Centers) as described in publications and other documents related to the field; The operation of a DSL system having an SMC) and / or a Dynamic Spectrum Management Center (DSM Center) or other suitable controller and / or entity including a computer system will be described. Where the term “controller” is used herein, it is intended to mean any or all of those control means, or any other suitable control means. A controller can be a single unit or a combination of components, which can be a computer-implemented system, a device, or a combination of devices that perform the functions described below.

  As will be appreciated by those skilled in the art, after reading this disclosure, embodiments of the present invention can be adapted to work with various DSL and other communication systems known to those skilled in the art. A dynamic spectrum manager or other controller that manages a communication system using one or more embodiments of the present invention is a service provider and / or operator (which may be a CLEC, ILEC or other service provider). Or it can be a party that is partially or completely independent of the system operator.

In general, in communication systems, if more parameters are monitored and tunable rather than statically set, performance can often improve dramatically (e.g., Higher data rates can be realized, more users can be serviced, less power can be consumed, etc.). That is, if system settings are set adaptively as a function of line or channel performance history, adaptive changes to system operation can improve data rates and other services for users. As an example, there are currently no systems that dynamically monitor a number of parameters, characteristics, etc., and help operators and users optimize DSL services. Some operators have built a basic form of collecting DSL line data, and
-Increasing the usable data rate after initial service introduction, and / or-observing the bit error rate of the circuit over time, until an acceptable functioning rate is observed (called provisioning) Attempting to determine if re-provisioning at a low data rate is required.

  Specifically, in those systems, the rules that increase or decrease the data rate are most often over-simplified fixed functions of only a few input parameters. Accept and analyze more inputs, essentially becoming a dynamic function consisting of several parameters based on the observation and processing of many other observed parameters and a history of line performance The system according to this embodiment constitutes a significant improvement component in this field.

  In order to reduce various types of performance problems, including crosstalk, many communication systems limit the power that may be used by transmitters that transmit data within a given system. The transmission system margin exceeds the minimum power required to achieve the desired performance (eg, system bit error rate threshold, or BER), a level of transmit power (typically expressed in dB). It is. The basic goal is to use enough power to overcome and / or correct for noise-induced and interference-induced errors and to reduce potential problems caused by excessive levels of transmit power. Is to minimize the power required for However, in many cases, equipment manufacturers, system operators, etc., to provide a high data rate and to take an easier approach to addressing potential issues such as crosstalk, Use such excess power (leading to excess margin).

  The present invention uses information regarding line characteristics (eg, operational data) in power adaptation systems and procedures to more carefully assess acceptable problems / interference avoidance and data rates. This, after analyzing the available information and / or operational data, trains the modem to operate at a power transmission level (ie, margin) that will provide sufficient power for acceptable data transmission. And setting and minimizing the negative effects that one user may have on the other. More specifically, embodiments of the present invention generate margin related parameters for at least one modem in a modem pair using one or more such margin related parameters to allow a modem to determine a predetermined margin target. Can help you meet.

  FIG. 1 shows G.K. Fig. 2 illustrates a reference model system in accordance with the 997.1 standard (also referred to as G.ploam) with which embodiments of the present invention can be used. This model includes ADSL1 (G.992.1), ADSL Lite (G.992.2), ADSL2 (G.992.3), ADSL2 Lite G. 992.4, ADSL2 + (G.992.5) and G.992.4. 993. VDSL standard revealed by x. Applies to all ADSL systems that meet various standards that may or may not include a splitter, such as the 991.1 and G991.2 SHDSL standards, all with and without connections. This model is well known to those skilled in the art.

  G. The 997.1 standard is G. A clear EOC (embedded operation channel) as defined in U.S. 992. Define physical layer management specifications for ADSL transmission systems based on the use of indicator bits and EOC messages specified in the x standard. G. 997.1 defines network management element content for configuration, fault and performance management. In performing those functions, the system uses various operational data (including performance data) that is available at the access node (AN).

  In FIG. 1, a user terminal device 110 (sometimes referred to as a subscriber premises equipment or CPE) is coupled to a home network, which is coupled to a network termination (NT) unit 120. The NT 120 includes an ATU-R 122 (eg, a transceiver defined by one of the ADSL standards) or some other suitable network terminated modem, transceiver or other communication unit. The NT 120 also includes a management entity (ME) 124. The ME 124 may be any suitable hardware device such as a microprocessor, microcontroller or circuit state machine in firmware or hardware operable as required by any applicable standards and / or other standards. it can. The ME 124 particularly collects operation data and accumulates the data in the MIB. The MIB is a database of information stored by each MIB, and the MIB is a network such as SNMP (Simple Network Management Protocol). Management protocol, can be accessed via the management protocol used to collect information from the network device and supply it to the administrator console / program, or via TL1 commands, where TL1 is between telecommunications network elements Is a command language with a long history used to program responses and commands.

  Each ATU-R in one system is coupled to an ATU-C or other central location in the CO. In FIG. 1, the ATU-C 142 is arranged in an access node (AN) 140 in the CO 146. The ME 144 similarly holds an MIB consisting of operational data related to the ATU-C 142. AN 140 may be coupled to a broadband network 170 or other network, as will be appreciated by those skilled in the art. ATU-R 122 and ATU-C 142 are coupled together by line 130, which in the case of ADSL is typically a telephone twisted pair that also mediates other communication services.

  Some of the interfaces shown in FIG. 1 are used to determine and collect operational data. The Q interface 155 forms an interface between the operator's network management system (NMS) 150 and the ME 144 in the AN 140. G. All parameters specified in the 997.1 standard are applied at the Q interface 155. Near-end parameters supported by ME 144 are obtained from ATU-C 142, while far-end parameters from ATU-R 122 can be obtained by either of the two interfaces across the U interface. Indicator bits and EOC messages transmitted using the embedded channel 132 and generated at the PMD layer can be used to generate the required ATU-R 122 parameters at the ME 144. Alternatively, an operation, management and maintenance (OAM) channel and appropriate protocol can be used to retrieve parameters from the ATU-R 122 when required by the ME 144. Similarly, the far end parameters from ATU-C 142 can be obtained by either of the two interfaces across the U interface. Indicator bits and EOC messages generated at the PMD layer can be used to generate the required ATU-C 142 parameters at the ME 122 of the NT 120. Alternatively, the OAM channel and appropriate protocol can be used to retrieve parameters from the ATU-C 142 when requested by the ME 124.

  The U interface (which is essentially line 130) has two management interfaces, one at ATU-C 142 (UC interface 157) and one at ATU-R 122 (U- R interface 158). Interface 157 generates ATU-C near-end parameters for ATU-R 122 for searching across U interface 130. Similarly, interface 158 generates ATU-R near-end parameters for ATU-C 142 for searching across U interface 130. These applied parameters can depend on the transceiver standard used (eg G.992.1 or G.992.2). G. The 997.1 standard defines an arbitrary OAM communication channel that crosses the U interface. If this channel is implemented, the ATU-C and ATU-R pair may use the channel to send physical layer OAM messages. Thus, the transceivers 122, 142 of such a system share various operational data stored in their respective MIBs.

  As will be appreciated by those skilled in the art, at least some of the parameters described in these documents can be used with embodiments of the present invention. Also, at least some of the system descriptions are equally applicable to embodiments of the present invention. Various types of operational data obtained from DSL NMS can be found therein, and others can be known by those skilled in the art.

  In a typical topology of a DSL plant where multiple transceiver pairs are operating and / or available, a portion of each subscriber line is placed alongside other users' lines in a multi-pair binder (or bundle). Has been. Behind the pedestal very close to the customer premises equipment (CPE), the circuit takes the form of a drop wire and exits the bundle. As such, the subscriber line traverses two different environments. Part of the line can be placed inside the binder, in which case the line is subject to crosstalk, although it may be shielded from external electromagnetic interference. Behind the pedestal, the drop wire may not be affected by crosstalk because it is far from other active pairs for most drops, but transmission is not due to the drop wire being shielded There is a possibility that the function is remarkably impaired by electromagnetic interference. Most drops have 2 to 8 twisted pairs in them, and in the multiservice situation (single service multiplexing and demultiplexing) to the home or bonding of those lines, an additional significant amount Crosstalk can occur between those lines in the drop segment.

A general and exemplary DSL deployment scenario in which embodiments of the present invention can be used is shown in FIG. All subscriber lines of the total number (L + M) of users 291 and 292 pass through at least one common binder. Although the lines of FIG. 2 are shown with approximately the same length, the lines of a given system can vary in length, and in some cases, the length can vary significantly. Each user is connected to the central offices 210 and 220 via a dedicated line. However, each subscriber line may pass through a different environment and medium. In FIG. 2, L users 291 are connected to CO210 using a combination of optical fiber 213 and twisted pair copper wire 217, which is typically, F T TCab (Fiber to the Cabinet) or Fiber-to-the・ It is called a curve. The signal from the transceiver 211 of the CO 210 includes signals converted by the optical line terminal 212 and the optical network terminal 215 in the CO 210 and the optical network unit (ONU) 218, which are also called remote terminals (RT). The modem 216 of the ONU 218 functions as a transceiver for signals between the ONU 218 and the user 291.

The remaining M lines 227 of the M users 292 are only copper twisted pairs, which is a scenario called FTTEx (Fiber to the Exchange). Where possible and economically feasible, F T TCab is preferred over FTTEx because the length of the copper part of the subscriber line is shortened and, as a result, the realizable rate increases. . The presence of F T TCAB line, can cause problems for FTTEx line. In addition, F T TCab is predicted that in the future, become increasingly popular topology. This type of topology can result in significant crosstalk and can mean that different user lines have different data transmission and performance capabilities depending on the particular environment in which the line operates. There is. The topology may allow fiber feed “cabinet” lines and switched lines to be mixed in the same binder. Users L + 1 to L + M can be remote terminals (instead of CO), and users 1 to L can be closer to the user and can be a line terminal device or some other fiber feed terminal. Services can also be provided by the device (ie, two fiber feed terminals, one closer to the user than the other).

  As can be seen from FIG. 2, the line from the CO 220 to the user 292 shares the binder 222 and the binder is not used for the line between the CO 210 and the user 291. The other binder 240 is common to all of the lines to / from CO 210 and CO 220 and their respective users 291, 292.

  According to one embodiment of the present invention shown in FIG. 3, the margin and power analysis analyzer 300 optimizes the use of a DSL system by a user and / or one or more system operators or providers or other It can be part of an independent entity that monitors as a controller 310 (eg, a dynamic spectrum manager or a dynamic spectrum management center) that assists in controlling the method. (The dynamic spectrum manager may also be referred to as a dynamic spectrum management center, DSM center, system maintenance center or SMC.) In some embodiments, the controller 310 is operated by an ILEC or CLEC from a CO or other location. Can be activated by a DSL line. In other embodiments, such as the example in FIG. 12, the “high performance” modem unit 1200 is integrated with the modem 1210 (eg, processor and memory) at a user location, central office, or other single location. Controller 800 '. As can be seen from the dotted line 346 in FIG. 3, the controller 310 may be in the CO 146 or may be part of the CO, or may be independent of the CO 146 and within the system. It may be some party that operates. The controller 310 can also be connected to and / or control multiple COs. Similarly, components of controller 310 may or may not be in the same location and / or in the same device, and / or may be accessed instead by controllers at different locations.

  In the exemplary system of FIG. 3, analyzer 300 includes collection means 320 (which can also perform monitoring if desired) and analysis means 340. As can be seen in FIG. 3, the collection and / or monitoring means 320 can be coupled to sources internal to the DSL system, such as the NMS 150, the ME 144 in the AN 140 and / or the MIB 148 managed by the ME 144, and the source Data can be collected via and from sources. Data may also be collected from an external source by means 320 via broadband network 170 (eg, by the TCP / IP protocol or other means outside of a typical internal data communication system within a given DSL system). it can. For example, the controller can collect operational data from the ATU-R over the Internet or from the ATU-C over the Internet if EMS is inconvenient or bandwidth limited. The operational data can also be collected from the service provider's NMS, which can itself be collected from various sources.

  Analysis means 340 and / or monitoring / collection means 320 may also be coupled to margin performance or historical sources 345, such as a database or memory that may or may not be part of analyzer 300 or controller 310. The connection of one or more analyzers allows the analyzer 300 to collect operational data. Data can be collected once (eg, during a single transceiver training) or over time. In some cases, the monitoring means 320 collects data on a regular basis, but the means can also collect data on demand or on some other non-periodic basis so that the analyzer 300 can collect data. The user and line data can be updated as necessary.

  The analysis means 340 analyzes the data supplied to the analysis means to determine whether it is necessary to send an instruction to one or more modems to help the modem meet a predetermined margin target. Is possible. The analyzing means 340 of the analyzer 300 is coupled to the command signal generating means 350 in the controller 310. The signal generator 350 is configured to receive a margin related parameter value generated by the analysis means 340 for use by the modem, where the margin related parameter value is based on the operational data, and Support at least one modem to meet the margin target, and other command signals (eg requested or required MAXNOMPSD value, PSDMASK setting or CARMASK, MAXSNRM, MINSNRM, TSNRM, MAXNOMATP, MAXRXPPWR, etc. Or either a rate adaptation margin or a timer) is calculated to be transmitted to a user in a communication system (eg, an ADSL transceiver such as ATU-C). As represented by the dotted line 347, the command signal generating means 350 may or may not be part of the analyzer 300 and / or implemented in the same hardware, such as a computer system. The command signal generator 350 constitutes a means for adjusting one or more margin related parameter values in the modem pair.

As will be appreciated by those skilled in the art, the controller is a completely independent entity (ie, not owned and / or operated by the company that owns and / or operates the line in the CO) In some cases, most DSL system configuration and operation information may not be available. Even when a CLEC or ILEC operates and / or functions as the controller 310, most of this data is not known. Various methods can be used to evaluate the necessary data and / or information. One example of such a method is the “DSL System Evaluation and Filed on April 2, 2004, owned by Adaptive Spectrum and Signal Alignment, Inc.”, which is filed by Adaptive Spectrum and Signal Alignment, Inc. parameter recommended Ru can be found in U.S. Patent application Serial No. 10 / 817,128 of (DSL SYSTEM ESTIMATION AND pARAMETER rECOMMENDATION) of the title.

  In some embodiments of the invention, the analyzer 300 can be implemented on a computer such as a PC, workstation, etc. (one example is disclosed in FIG. 8). The collecting means 320, the analyzing means 340 and / or the instruction signal generating means 350 can be software modules, hardware modules or a combination of both, as will be appreciated by those skilled in the art. All of these components can be in the same computer system, for example, or they can be in different devices. In the case of managing a large number of lines, a database may be introduced, and the database can be used to manage the capacity of the lines and the data generated by the controller.

  The configuration of FIG. 3 can be used to implement a power adaptation system and method according to embodiments of the present invention. As can be seen in FIG. 4A, a power adaptive system, such as a system included in the present invention, reduces and / or minimizes power consumption while maintaining a target data rate (eg, minimum data rate) and target noise margin. To do. The rate adaptation system and method shown in FIG. 4B use all available power (generally total transmit PSD at a fixed level) to maximize the data rate and maintain the target margin level. The margin adaptation system and method shown in FIG. 4C again uses all available power to maximize the margin and maintain a constant data rate. The current arrangement of ADSL follows the margin adaptation method of FIG. 4C, which in many cases is detrimental to the user and operator. If excessive power is used and data rates are unnecessarily boosted or excessive margin levels are generated, crosstalk and other problems can occur to users and operators. Unlike excessive power systems, a power adaptation method such as the present invention can achieve a sufficient margin to ensure reliable data rate, minimal power consumption and reliable error and interference avoidance.

  Operation in the ADSL field often requires that modems often conform to many new and changing DSL standards, but all of the various quantities, rules and guidelines are applied to different manufacturers ( And in fact, it reminded me to interpret in different ways (for different generations and versions of the same manufacturer's hardware and software). Also, the various interconnect tests, including those close to release, do not adequately accommodate all the various possible configurations and power levels, leaving a wide range of uncertainties regarding actual field conditions. Using the present invention, the controller can reduce, increase, and / or maintain power and / or PSD levels to ensure robust operational guidelines to avoid problems such as excessive crosstalk between modems. And impose and implement implementation. In addition, the controller allows the rate / performance / price combination to provide DSL service to the user in a binder with varying crosstalk and the degree of user topology, resulting in maximum service performance and / or revenue. Can be experimented with various settings that are not generally imposed by predicted operation or during operation of a standard-compliant product to ascertain the cause / effect and time variation of the DSL environment.

  The time-varying information and method corresponds to an ADSL2 mode known as dynamic rate adaptation. G. for this purpose. The specific parameters in 997.1 are known as RA-USNRMus / ds and RA-DSNRMus / ds (rate adaptive upshift / downshift signal-to-noise-ration margin, upstream or downstream rate). Allows setting of a margin target that must be realized before descending. RA-USNRMus is a level that is compared with the calculated margin of the modem. If the calculated margin exceeds USNRMus for RA-UTIME period or longer, the data rate can be increased without retraining in ADSL2. After the rate increases, the margin becomes smaller than before the rate increases. Here, if the calculated margin is smaller than USNRMus, the modem's calculated margin is compared with RA-DSNRMus, and if it is greater than this value, the modem continues the same data rate. Here, or at any point in time, if the margin is below RA-DSNRMus for a period exceeding RA-DTIMEus, the modem data rate is reduced until the margin again exceeds RA-DSNRMus. Rate adaptation stops and there will always be a maximum rate at which MAXSNRM is applied. The margin target must be maintained for a period specified by the DSM center by other control parameters RA-UTIMEus / ds or RADTIMEus / ds (rate adaptive upshift / downshift time, upstream or downstream).

  In general, as shown in the embodiment of FIG. 7, the controller collects operational data (typically for the DSL modem pair of interest) at 710. Operational data includes historical margin performance of the DSL system, historical performance data (such as previously measured and known margin levels for the modem pair and other performance related information), current performance data for the DSL modem, retraining count data. , Other data related to modem training, or error data.

  Data can be collected using the internal communication system of the DSL system and / or using external communication (eg, the Internet). The operational data may include information regarding one or more modem operational parameter values collected at 720 that are used by or set by the modem pair.

  At 730, the controller analyzes the operational data to determine which margin-related parameter values can help the modem pair meet a margin target or improve the performance of the modem pair. The controller then generates a margin related parameter value at 740. The margin related parameter values may be for modem operating parameters as considered by the controller, or may be different margin related parameters. At 750, the controller generates a command signal representing a margin related parameter value and sends the signal to at least one modem of the modem pair, thereby allowing the modem pair to be trained or in normal operation, depending on the situation. To use a margin related parameter value for the use of.

  The controller may update the operation of the modem pair by performing such an analysis more than once, as indicated by the dotted arrows in FIG. 7, or it may be updated at a predetermined time, such as immediately before modem training. May be performed. As discussed in detail below, the above parameters that the controller operates with and the operational data available to the controller vary depending on the type of DSL system in which the modem pair operates. Again, when analyzing modem margin performance, the modem operating parameters used by the controller may or may not be the same parameters that the margin-related parameter values are generated and sent to the modem. . While not limited to such types, embodiments of the present invention are useful for supporting modems that utilize ADSL1, ADSL2, ADSL2 + and / or VDSL. The use of a controller can help ensure that a standards-compliant modem remains compliant. Also, embodiments of the present invention can improve the performance of one or more DSL lines by taking into account operational data such as the effects of crosstalk and other information that may have a detrimental effect on DSL performance. Can be used to improve.

  The basic idea is that the spectral level, power, spectral shape, etc. can all be changed according to the reported margin history / distribution. In other words, after evaluating the data on the previous performance of the modem pair and knowing the margin-related parameters of one or more modem pairs, the controller or the like will determine whether the modem is 1 or whatever the modem pair imposes according to the standard. It can be suggested or enforced to employ an operational value that assists in meeting one or more margin targets.

  In some embodiments of the present invention, the controller coupled to the ATU-C side of the modem pair imposes different MAXNOMPSD levels (eg, in ADSL2 systems, by setting and / or changing the MAXSNRM parameter). Or by setting PSDMASK on the ADSL2 + modem or some or all of them, or other parameters mentioned above, such as CARMASK, MAXSNRM, TSNRM, MINSNRM, RA-margin / timer, etc. Dynamic control of margin setting and adjustment for each line (with some combinations). Such dynamic margin setting by imposing a lower mask is not part of any standard. Even those attempting to comply with margin and power rules can be limited by the range of PCBs allowed, or early training and (possibly prior to other modems and / or line ends). This can be limited by not having an available history of information from line usage (with other users who may have existed). Thus, in another embodiment, the controller may look at the history of reported margin measurements where the line exceeds the desired margin target, thereby allowing more of during or before training, depending on the mechanism described above. A low PSD level may be imposed. This is because users and operators do not actually know the expected performance level, and therefore did not want to unnecessarily weaken the modem if they could experience high noise during operation. In the previous system, it was not done. Similarly, if for some reason the modem is not using enough power and / or margin and is exposed to excessive noise and error problems, the controller can allow the modem to perform better. In order to be able to instruct to use a higher PSD level during training or operation.

  As mentioned above, in some systems, “seeding” the training process with historically measured and / or known margins so that appropriate power reduction is performed during training. Is preferred. The controller can store performance history or can access performance history, so the controller continually improves evaluation and judgment regarding PSD or other margin related parameters so that the modem Allows the modem to be instructed to be used when it is reset or retrained (this can be forced or recommended where appropriate). For example, the service provider or controller waits for the above line to become inactive, e.g. waiting for an ATM cell count or other user information passing measurement to know when the line is active or not, and then resets it. Thus, a new PSD can be used in a way that the user does not notice. In other situations, the service provider may simply retrain when the system is unlikely to be used (eg, midnight). In some embodiments, the controller uses this historical information to allow one or both of the modem pairs (eg, ATU-C) to use available PCB values or other adjustments (eg, by ATU-R). Which initial PSD level should be used so that (-14.5 dBm drop) has an opportunity to meet the margin specification.

  In some embodiments of the present invention, programming is based on prior use or training. Prior use may be more important in some cases. The second pass-through training that can also be used is essentially a pause for the modem vendor itself, especially in the case of downstream transmissions to the DSLAM vendor, in which case the modem essentially takes the current training. Training can be started a second time with a different lower NOMPSD that stops and makes the margin less than MAXSNRM.

Some embodiments of the present invention combine to observe ADDNMR limits by initializing PSDMASK settings using the prior knowledge. Optimum spectrum management (OSM), known to those skilled in the art, is from a level 2 adjusted spectrum management by a dynamic spectrum manager, DSM center or other controller that has been studied and exceeds the already significant benefits of theoretical repeated water injection. It shows some benefit, which is addressed in earlier systems. “Level 2” means that a controller, such as a DSM center, can adjust spectral levels together (eg, based on recognized crosstalk between two or more lines). Level 1 means setting the spectrum based only on observations from the same single line. Level 0 means that there is no ability to perform DSM. Additional information regarding OSM is preferred draft TIE1.4 / 2003/325, including TIE1.4 / 2004/459 and TIE1.4 / 2004/460, it is found in the various contributions to TIE1.4 Working Group ANSI it can. However, the central adjustment required for OSM makes it difficult to realize in an actual system, mainly due to the need to control the spectrum. In an embodiment of the present invention, a new DSM report for ANSI TIE 1.4 and G.2 for VDSL2. The use of 997.1 PSDMASK is essentially the same performance as OSM by simply setting a small number of flat PSD masks on different segments of the frequency used by DSL modems. Can be realized. The level of those bands can be increased or decreased until the desired combination of data rates is realized between users, and the user can still apply bit-to-tone or bit-load in the normal way and apply to every single tone. Monitor for specific PSDMASK limits possible. The reported margin can correspond to the worst case margin, and MAXSNRM is generally only applied to the tone with the smallest margin. Manufacturers may not use the best loading or bit-swapping algorithms that lead to imposed PSDMASK variations and interpretations. Therefore, apply the preferred band (or “PREFBAND”) bit or indication that discontinuous water loading or approximation to it is preferred, and the MAXSNRM parameter to the margin found on all tones (not just the worst tone) It can be sent to the EM of the modem pair to signal that it should. This PREFBAND instruction is part of the present invention.

  As described above, in the case of an ADSL1 system, the MAXNOMPSD value is generally set by the operator. However, using one embodiment of the present invention, one example of which is shown in FIG. 5A, the controller 510 provides margin related parameter values (eg, MAXNOMPSD values) to the ATU-C 530. Controller 510 may send instructions or may communicate with the system using an element management system that can receive MAXNOMPSD or PSDMASK from NMS and / or controller 510. As will be appreciated by those skilled in the art, the command signal (eg, from command signal generator 350) is sent to ATU-C 530 by NMS, element management system, email, ftp, or in some other suitable manner. Can send. The controller may also supply a MAXNOMATP value in some embodiments in addition to or instead of MAXNOMPSD. In some cases, an ADSL1 CARMASK processing procedure (a simple for each tone that is standardized and allowed for operator specifications in ADSL1 is used to turn off the carrier in a band that causes excessive crosstalk to other DSL systems. Can also be used instead of or in addition to MAXNOMATP / PSD.

  In an embodiment of the invention applicable to ADSL1, the controller 510 may be configured to operate from the operational data (eg, MIB 525 or historical data module 520) collected by the controller for training of the transceiver that leads to the appropriate margin during SHOWTIME operation. Margin-related parameter values (for example, MAXNOMPSD value) used by the modem. (NOMPSD is selected by the transmitter, not the setting managed by the MIB, but NOMPSD must be smaller than MAXNOMPSD, and if the fieldwork can implement the value, any MAXNOMPSD This is for transmitters that set NOMPSD to the same level of: Some MAXNOMPSD supplied to RT that can only implement -44 dBm / Hz, for example, -40 dBm / Hz is more accurately than NOMPSD The history / library 520 obtains data from the operator's MIB 525 and any other available source of relevant data regarding system performance. The controller 510 supplies the MAXNOMPSD value or other margin related parameter value to the ATU-C 530. The MAXNOMPSD value provided by the controller can be calculated so that the system achieves a TSNRM / TARSNRM margin value, a MAXSNRM margin value, a margin value between the two, or other desired margin targets. The controller can perform line performance under various different types of noise situations that have occurred, are currently occurring, or can occur, especially under simulated conditions in which other adjacent lines are also programming PSDs. You may choose to inspect or estimate.

  Since modem 510 often uses MAXNOMPSD as its NOMPSD value, the new MAXNOMPSD value supplied by controller 510 is likely to be the NOMPSD value used by ATU-C modem 530. Even if the MAXNOMPSD value supplied by the controller is not selected by the ATU-C as the NOMPSD value, the NOMPSD value will not be higher than the supplied MAXNOMPSD value and excess margin will be avoided during normal SHOWTIME operation. .

  Thereby, the controller imposes an upper limit on NOMPSD, but the controller cannot impose a NOMPSD value directly. Thus, if -52 is the desired NOMPSD value, the controller sets MAXNOMPSD to -52. Since NOMPSD cannot be higher than MAXNOMPSD, setting MAXNOMPSD to -52 limits the value of NOMPSD, the value also limits the level of the resulting margin, and the use of excess margin To avoid. Generally, MAXNOMPSD is the NOMPSD in the very fast transceiver training part of ADSL1 training, but this is up to the vendor, so the controller indirectly imposes the NOMPSD value used by the ATU-C. Can do.

  As can be seen in FIG. 5A, after the ATU-C 530 receives its MAXNOMPSD value (and resets as needed), the ATU-C 530 may “set by the controller” or “manipulate by the controller”. NOMPSD is used to measure upstream power transmission from ATU-R 540 and thereby evaluate the line length. Based on this evaluation, the ATU-C 530 calculates any PCB power drop (eg, according to Annex A of the ADSL1 standard), informs the ATU-R 540 of this value, and proceeds to step 550 of FIG. 5A. As shown, PEFPSD = NOMPSD-PSD is set. Using embodiments of the present invention, PCB values are less likely to be compliant with margin requirements due to consideration and use of historical data 520 stored by controller 510 for the line on which modems 530, 540 operate.

The ATU-R 540 then calculates the margin, bit (b _i ), and gain (g _i ) after transceiver training and channel analysis using the REFPSD value. The value of _{g i} that are allowed under the ADSL1 is -14.5dB~ + 2.5dB. As can be seen in step 560 of FIG. 5A, the final PSD value for tone i is PSD _i = NOMPSD−PCB + g _i according to the MAXNOMPSD limit initially given by controller 510. (In some embodiments of the invention, the gi value may be a margin related parameter value and indirectly by the controller via a command signal from the controller to the receiver that commands it to a lower gain. Thus, if NOMPSD is MAXNOMPSD, then the PCB is 2 dB and the g _i for the large population of adjacent tones is +2.5 dB, so that the final for all tones Even though the calculated PSD value is 0.5 dB greater than MAXNOMPSD (which is unlikely to occur in light of analysis of historical data and selection of a MAXNOMPSD controller based on previous operating characteristics), MAXNOMPSD Will be limited to an average value due to the limitation of MAXNOMPSD for . In theory, this can be applied to each tone. However, ADSL1 allows MAXNOMPSD to exceed 2.5 dB for individual tones, but on average for tones, MAXNOMPSD must be monitored. When the combination of PCB and g _i is not positive, the PSD _i value is NOMPSD or less.

  Thus, the NOMPSD affected by the controller “seeds” this entire process, and the controller 510 deducts the resulting margin and excess margin from the ADSL1 standard (or any other imposed restrictions). This makes it possible to comply with the MAXSNRM level supplied by the MIB specified in (1). Finally, the adjustment can be performed during SHOWTIME using the gain swapping capability of ADSL1.

  There is no MIB parameter that tells the controller 510 directly which NOMPSD values are in ADSL1. The controller may base its recommendations / instructions on a training sequence that has just been completed (in this case, the modem pair directly enters the retraining procedure) or other historical data that the controller can access. .

  For upstream power reduction, the ATU-R 540 starts by sending a test signal to the ATU-C 530 at a preselected PSD value, eg, -38 dBm / Hz, as shown in the example of FIG. 5B. The ATU-C 530 measures the line attenuation, evaluates the line length, and transmits the information to the ATU-R 540. ATU-C 530 also calculates margins, bits, and gains for its own operation. The second transmission of ATU-R 540 still retains its original PSD value. After receiving the second transmission from the ATU-R 540, the ATU-C 530 calculates its gain and if the initial PSD value of -38 is used by the ATU-R 540, the final ATU-R 540 A power drop of about 14.5 dB can be commanded so that the PSD is between -52.5 and -38.

  While upstream and downstream training / power reduction are separate events, one embodiment of the present invention lowers the upstream PSD under controller control if the previous margin is high. Upstream crosstalk is generally not bad in DSL, but a strong upstream signal can cause more echoes in the user's modem to leak into the downstream signal. By reducing this echo by a lower non-standard upstream PSD, the downstream performance often does so in the presence of bridge taps, so if the echo is superior to other noise, it is a few dB (case Can be lifted up to about 10 dB). With this embodiment, the modem will outperform the current modem over long lines, in which case this local echo in the user's modem will most likely be the dominant range-limited. I think it will have an impact.

  After the modem enters SHOWTIME, further adjustments to gain can be performed using gain swapping.

  In ADSL1 and other systems, a simple option according to one embodiment of the present invention is to measure the margin immediately before SHOWTIME. If the measured margin is higher than a predetermined limit just before the modem enters SHOWTIME (eg, 16 dB MAXSNRM), training resumes and the modem uses the allowed cutback value to Retrain on pass. Such an implementation cannot be controlled by the DSM center in most situations, but instead in a proprietary mode within the modem itself, for example by a method according to an embodiment of the invention in a software module etc. Can be executed.

  Embodiments of the present invention implemented under the ADSL2 standard, examples of which are shown in FIGS. 6A and 6B, again, initialization process, handshake, channel discovery, transceiver training, channel analysis and modem Seeds exchange using margin-related parameter values supplied by one or more controllers for.

  Due to the range of available PCB values (0, 1, ..., 40 dB) and the fact that either ATU-R or ATU-C can command cutback, ADSL2 standardizes a wide range of power reductions. Having a mechanism, which can be used to instruct the transmitter to reduce the initial PSD if a high margin has been observed first and continuously. The value of the transmitter PCB can be calculated starting with the MAXNOMPSD parameter supplied by the controller. Accordingly, embodiments of the present invention assist in setting MAXNOMPSD and / or PCB / tssi with a proprietary level of PSD lower than 60 dBm / Hz using past history.

  As mentioned above, the transmitter can be reduced by PCB when instructed to do so by a DSM center, operator or other controller. In ADSL2, the transmitter can impose a PCB because the modem must use the larger of the two PCBs required by the transmitter and receiver. Typically, compliance with operator-supplied MAXNOMPSD is performed and the NOMPSD may already be below -40 dBm / Hz before the PCB is used early in the training. G. ADSL2 training The actual NOMPSD sent in one of the fastest messages sent between the transmitter and receiver in what is known as the hs part occurs before the PCB notification. And the downstream transmitter can impose a more substantial PCB for several reasons (eg, if the transmitter wants to observe MAXSNRM reliably because the upstream signal is large, Or because the transmitter is ordered by the manufacturer's proprietary mode to use the PCB by the operator). However, the external MIB for very low PSD is not in the ADSL2 standard (MAXNOMPSD ≧ −60 in ADSL2).

Using the present invention, adjustments up to 40 dB are required based on the high margin reported earlier in the line or line history. The transmitter allows the controller (eg, DSM center) to specify that the transmitter requires a PSD lower than -60 dBm / Hz (which is not strictly necessary in the ADSL2 standard). Since MAXNOMPSD cannot be lower than −60 dBm / Hz in ADSL2, the controller (eg, DSM center or operator) can force additional reduction by PCB or tssi parameters. The tssi parameter exists in ADSL2, but cannot be specified by the controller or operator in ADSL2 (in ADSL2, only the modem manufacturer can set tssi). (ADSL2 + allows tssi to be operator specified by the PSDMASK MIB parameter. In ADSL2 +, the controller, operator or DSM center can enforce the tssi value by means of an MIB parameter called PSDMASK. PSDMASK MIB. The parameter is not in ADSL2, but instead the manufacturer can set the tssi value as desired.) In summary, the PCB is actually under the control of the transmitter. The controller (eg, DSM center or operator) is very low for transmitting PSD if the ADSL2 modem allows the operator or DSM center to set the PSD below -60 in a proprietary mode. PCB can be indirectly specified by setting. (Some ADSL1 DSLAMs have software upgrades that exceed the scope of standard G.992.1, and the MAXNOMPSD value can be set as low as -80 dBm / Hz by the controller / DSM center. In particular, the line does not know this has been done, but may still appear long to the ATU-R that is still functioning well. (It can be set by the method of complying with the standard using tssi in ADSL2 + by PSDMASK specified by MIB of the
In the upstream direction only, the PCB transmitted to the transmitter by the receiver (ATU-C) can set the upstream received power in the ADSL2 by the operator or the controller (for example, the DSM center). It can also be requested to be less than the value known as possible MAXRXPWR. Therefore, by using MAXRXPWR, the controller can cause the PCB to be implemented via a DSM center command or the like. However, downstream requires the use of proprietary functions to force below -60 (using either PCB or tssi). The present invention either sets the level using tssi, or implements a DSM high performance transmitter that avoids the need for controller commands and sets the PSD level internally using PCB, Or both.

  6A and 6B, the ADSL2 standard allows the receiving modem to command 0, 1, 2,..., 40 dB power reduction (PCB) as part of training. This additional usable maximum value of 40 dB does not exist in the transmission system of ADSL1. The transmitter may also decide to reduce power and inform the receiver that it is operating so that either modem can specify a PCB value. Of particular importance is that the ATU-R can do this for downstream transmissions and the ATU-C can do this for upstream transmissions.

  In the embodiment of FIG. 6A, the controller 610 sends a margin related parameter value (eg, MAXNOMPSD value) to the modem pair at step 645 (eg, sends the parameter value to a single modem such as ATU-C 630). To start). Also, the ADSL2 and “G.ploom (G.997.1)” standards are defined by MAXNOMATTP, which would have the same effect as the MAXNOMPSD parameter, so that an operator or controller (such as a DSM center) is external The power can be limited from. According to ADSL2, NOMPSD must be -60 to -40 for each tone i for which CARMASK is 1, allowing selective use of frequencies within the available band. . As with ADSL1, the NOMPSD value is typically set to MAXNOMPSD to allow the use of the maximum allowable margin.

  The controller 610 may examine a margin performance history 620 (such as a library, database, memory or computer module) that can obtain that information from any suitable source, eg, system evaluation or MIB 625, and the MIB The data is obtained from ATU-C 630, one or more management entities 644, or from some other source as will be appreciated by those skilled in the art. Using the collected operational data and possibly one or more modem operational parameters, the controller 610 analyzes the operational data to generate one or more margin related parameter values and implements the modified margin related parameter values. Can be sent to the modem pair. Margin related parameters (here, for example, the MAXNOMPSD value sent to ATU-C 630) will produce an appropriate margin level after some pre-operational transceiver training, etc. to help the modem meet one or more margin targets. To be selected and calculated.

  ATU-C 630 sends the starting NOMPSD value to ATU-R 640 during the handshaking phase of the training according to Article 8.13.2 of the ADSL2 standard. During the channel discovery phase 650 of pre-operation training, ATU-C 630 and ATU-R 640 each measure line performance and suggest PCB values. The maximum PCB value (ie, maximum power reduction) is taken by modem pair 630, 640, thereby setting REFPSD for each frequency in use, where REFPSD = NOMPSD-PCB, and It means that the lower of the two REFPSD values proposed by the two modems (transmitter and receiver) is used. In ADSL2, the NOMPSD level can be set to any level between −60 dBm / Hz and −38 dBm / Hz by the operator or controller 610 at intervals of 0.1 dB. Therefore, if the ADSL2 receiver accurately observes MAXSNR and inevitably orders a large value of PCB, a REFPSD level as low as -100 dBm / Hz may occur. This, of course, infers that the modem manufacturer has implemented the PCB correctly for MAXSNRM, but this parameter is often not implemented correctly in current modem interoperability / compliance tests. Not only are they inspected and measured directly.

  During transceiver training 655 and channel analysis 660, the system sets up its equalizer and echo canceller and measures the SNR in each of the downstream and upstream directions. The last stage before SHOWTIME is the exchange in step 665. During this last pre-operation phase, the ATU-R may command a further power adjustment from -14.5 to + 2.5 + EXTGI. The system then enters its normal SHOWTIME operation, during which MAXSNRM is monitored and adhered to, and during that time, further adjustments may be performed with allowed gain swapping.

In ADSL2, the same steady state allowable gain value of -14.5 dB to +2.5 dB found in ADSL1 is nominally used. However, in ADSL2, any re-interpretation of any sync symbol power level that is brought to the same level as the data symbols allows the full range of this gain from -14.5 dB to +2.5 dB to be implemented, whereas the ADSL1 modem typically -40 deviates from the initial training gain _{g i} level during swapping, -42, it is limited ... or -52 initial power level of + 2.5 dB or within -2.5 dB. The ADSL2 EXTGI parameter allows gain to be increased from 2.5 dB to +18.0 dB in addition to the level nominally used for ADSL2 swapping. EXTGI is determined and / or set outside the influence of the operator, controller and / or dynamic spectrum manager by the DSLAM manufacturer and communicated to the receiving modem during initialization. A high value of EXTGI may confuse various PSD levels, even if this EXTGI is large enough, it should not exceed MAXNOMPSD or PSDMASK (ADSL2 +). Some manufacturers ignore the mask and perform EXTGI with the mask ignored because of the confused intention specified by the large EXTGI value and the lower mask (ADSL2 does not allow this) Good.

  Some current modems and systems may not be able to set up correctly on a regular basis, and MAXSNRM limits may not be correctly implemented in ADSL2 modems. Thus, by providing the controller-specified MAXNOMPSD, the transmitter ATU-C side can impose a PCB (and / or tssi in ADSL2) in the initialization message that informs the downstream PCB value, and the margin ( It effectively enforces a reduction by an amount equal to the proposed downstream PCB (based on the past history of margin observations). Specifically, in one or more embodiments of the present invention, the controller 610 provides the ATU-C 630 with a MAXNOMPSD level that is intentionally less than −60 dBm / Hz to force a lower margin, And close the margin to the commanded MAXSNRM or monitor the commanded MAXSNRM (based on a history of margin activity where a single training of the line is not revealed to the receiver modem's PCB decision algorithm) be able to.

  Also, if the PCB values are exchanged, the exact level of power reduction required to satisfy MAXSNRM may not be known, meaning that the PCB values may be too conservative. Thus, again, in ADSL2, controller 610 monitors and / or examines margin history 620 to provide a PSD on the line that is -52 for ADSL1 and lower than MAXNOMPSD of -60 for ADSL2. May need to be imposed. As a result, this will be monitored by the modem by the PCB for REFPSD = NOMPSD−PCB <MAXNOMPSD used. In addition, ADSL2 and the favorable G. The 997.1 standard (depending on the applicable annexes, some of the annexes only allow -40) allows the operator or controller of the DSL system to reduce the REFPSD level to as low as -60 dBm / Hz, Alternatively, it is possible to impose a MAXNOMPSD parameter that is restricted to a high level of about -38 dBm / Hz.

  In some cases, MAXNOMPSD (in effect, NOMPSD) may be set as high as -34 dBm / Hz (which is not standard compliant). The controller can instruct DSLAM to increase, in those cases, for very long lines, where the current -40 dBm / Hz is only about 12 or 13 dBm of transmit power. (Long lines with low margins and low data rates will be the only ones that actually need full power). The actual PSD lower than −60 dBm / Hz needs to be specified by the PSDMASK parameter observed as the MIB control parameter in ADSL2 +. However, the controller may also cooperate with ATU-C or ATU- to warn entities to use PCB (or tssi) values that further reduce -60 in a standards compliant manner even in ADSL2. R may be notified. The MAXNOMPSD parameter is communicated to both sides of the DSL line before ADSL training begins (in accordance with ITU standard G.994.1) with a procedure called “handshake”. The PSDMASK parameter is the only MIB specified in ADSL2 + (the parameter is communicated at the transmitter's discretion in ADSL2, but not controlled or specified by the operator), but in other ways (eg, Can be communicated by a file transfer program ("ftp") or a simple email message over the Internet to an agreed IP address between the controller and the ATU-C or ATU-R.

  One skilled in the art will note that the ADSL2 + standard has all of the ADSL2 capabilities specified above for up to twice the spectrum of the transmitted DMT tone used or up to twice the number. Therefore, the power reduction descriptions and functions discussed with respect to ADSL2 systems also apply to ADSL2 + systems. In addition, some embodiments of the present invention related to ADSL2 + may allow an operator and / or controller to specify a potentially non-flat initial PSDMASK that specifies a number of flat spectral regions having a series of corner frequencies and power levels. It uses a concept known as a spectral toolbox that makes it possible to impose quantities. PSDMASK can be as low as -96 dBm / Hz and as high as -32 dBm / Hz. This capability may be ignored, misunderstood or not performed by the sending system, which was detrimental to such a system. Using the past history of the line, the controller can implement embodiments of the present invention using some of the features of ADSL2 +.

  In ADSL2 +, up to 32 breakpoints (about 8 bands) are allowed. The mechanism for breakpoint transfer at 994.1 initialization actually allows the specification of ADSL2 + PSDMASK levels for all 512 downstream tones and all 64 upstream tones, as needed. Thus, a controller such as a DSM center or operator can increase or decrease the various bands that begin at the start of training and continue through all other training and SHOWTIME bit / gain swapping by imposing PSDMASK. Although the PCB can still be used, PSDMASK is essentially a replacement for ADSL2's MAXNOMPSD (although MAXNOMPSD still exists and PSDMASK cannot exceed it). It becomes a receiver mechanism that implements even more MAXSNRM. Parameters equivalent to PSDMASK must also be implemented by direct specification upstream of the tssi parameter (at the same time downstream can either specify tssi directly or easier direct specification of PSDMASK Except G). Permitted upstream under 992.5. The PREFBAND bit in the DSM report Corresponds to the additional ambiguity of PSDMASK when monitoring MAXSNRM, ie (at the limit, eg 15 or less but some finite value) G. using a bit cap. The margin calculation at 992.5 is typically defined as “worst margin across all tones”. Thus, the loading algorithm can increase the margin to a very high level in the preferred (i.e. good or used) frequency band, and smaller margins in a somewhat suitable band with lower PSD. Can be limited to values. Since the worst margin occurs in a somewhat favorable band, it is reported and essentially disables the use of the MAXSNRM principle. Instead, if PREFBAND is on, it means that the margin of all tones must be less than MAXSNRM as well as the worst margin. Thus, the receiver may confuse the intent of PSDMASK used for bandwidth priority or for other reasons. As a result, imposing a PSDMASK is signaled by the operator to the modem with PREFBAND "on". Indicating “on” essentially prevents the vendor's proprietary loading method from overriding the intention of priority band processing intended by imposing PSDMASK with PREFBAND on.

  For example, downstream, the ATU-R loading algorithm can verify that the margin is 7 dB for one tone somewhere in the band and all others are 30 dB or more. Since this is the worst margin, the modem requires compliance with the margin target MAXSNRM. However, PSDMASK is set so that the margin for any tone does not exceed MAXSNRM. PREFBAND ON means that the modem manufacturer must enforce all compliance and cannot show exactly 7 dB for one tone, and 30 anywhere else, With PREFBAND “on”, the MAXSNRM on all tones must be monitored.

Normally, the low level in PSDMASK is used to specify the band that should be avoided and / or simply used. However, the flat low PSDMASK <60 dBm / Hz specified by the operator / DSM center forces NOMPSD to this specified level (which can be as low as 96 dBm / Hz) in the present invention, thereby The low initial NOMPSD level inevitably imposes a low margin (even before the modem does the same with the PCB). This very low PSD is exercised or enforced when specified by the controller, operator or DSM center, and even if some manufacturers' modems have margins and margins, even if the standard enforces better management. Predict that there may be insufficient jobs for spectrum management. Along with the available NMS communications, the controller can communicate PSDMASK as a margin related parameter value to the modem via the Internet (eg, email and / or ftp communications). PSDMASK, because did not come to the modem through the element management system, modem, adjusting PCB and / or ((-14.5) ~ (EXTGI + 2.5) in the limit) the _{g i} with respect PREFBAND value Did not have to run.

Thus, as required by the standard, as an example in the case of a DSL line that allows 18 dB of EXTGI and a minimum MAXNOMPSD of −60 dBm / Hz, the ATU-R has a g _i of −58 dBm / By setting -14.5 dB in the band initialized with Hz, the PSD can be adjusted to -72.5 dBm / Hz in the band. The same ATU-R can set a PSD of −40 dBm / Hz in another band by setting gi = + 18 dB in another band. The resulting bandwidth priority is 32.5 dB (the difference between the two levels). This is done without using PSDMASK directly in the MIB of the element management system, but the receiver implementing the loading is informed of the degree of PSDMASK or bandwidth priority over the Internet, and here all of the gi With ranges, you will know to realize the range even if the initial PSD setting for PSDMASK is not allowed. That is, the intelligent ATU-R essentially forces a standard compliant ADSL2 ATU-C modem (in this case, PSDMASK not to be used) to enforce bandwidth priority. This results in the use of the full range of available g _i, therefore naturally, tssi, if possible, a better implementation. However, tssi is not always available, as is the case with ADSL2. On the other hand, if PSDMASK is working with MIB of ADSL2 + modem, this procedure is not required and tssi can be used instead.

  Some manufacturers allow external designation of MAXNOMPSD down to -80 in their new ADSL1 and ADSL2, which is a (slightly) non-standard operating mode, Does not cause any harm, other than having to know to adjust the calculated attenuation based on some parameters like Hlog and -52 (or -60).

  There is a reference noise in the upstream PSD in the VDSL that can be used to force the PSD to a particular setting (referred to herein as REFNOISE). In some embodiments of the invention, the controller acts back to the PSDMASK imposed as a margin related parameter value modified from the initial REFNOISE value. That is, with some variations and free use of the reference noise PSD, the present invention can be used with VDSL systems and can achieve the same effects as ADSL.

“Bandwidth priority” is important for achieving gains in OSM, and is a synergy that continues the normal bit-swapping and gain-swapping procedures required for a practical system without loss of performance. It can be defined as band enhancement at no loading. The dynamic spectrum manager cannot expect to deal with and change the bit allocation fast enough and will certainly be slower than the modem's response itself, which is enforced by the OSM proponents. This is one of the reasons why we agree that the water injection done by the modem in or near the way (sometimes called "repetitive water injection") is good enough. Band priority is essentially the receiving modem, either by PSDMASK or “load with priority” ADSL2 + G. The “tss ” parameter in the 992.5 “Spectrum toolbox” commands to monitor PSDMASK in the band specified as the margin related parameter value, and generally allows priorities for several bands during water injection, Prevent loading to levels that can interfere with other DSLs.

Injection procedure of the theoretical, relative to the energy E _n nonnegative each of the tone

(NSC is the maximum number of tones). The gap Γ is a constant determined by code selection and a desired margin at a predetermined bit error rate in DSL. The channel attenuation for each frequency is specified by | H _n | ² and the noise energy for each tone is specified by σ _n ² , both measured during training (or their ratios are directly measured) and SHOWTIME. Updated during operation. This processing procedure is considered to be executed periodically or in conjunction with updates at each channel change.

  Theoretical irrigation procedures are well known in DSL and can be approximated in a number of ways, including various greedy algorithms for discrete integer bit restrictions (also called the Levin-Campello procedures) In this case, consecutive bits have a desired maximum bit rate limit less than the maximum specified margin (known as MAXSNRM in various DSL standards) and known as TARSRMM or TSNRM in various DSL standards. Until the minimum or target margin is reached. Frequency-dependent bit caps and frequency-dependent TSNRM [n] that augment existing systems are when communicating by ftp and / or email with a receiving modem implementing a loading algorithm. These are incapable of FEC adaptation and the system does not occur frequently but is forced to perform frequency dependent protection against large intermittent / impulse noise if it occurs and noise affects This is useful when the frequency range giving is known.

The loading algorithm calculates the number of bits b [n] for each tone and the gi “gain” (g _n ) factor for each tone. There are many variations of loading algorithms that are well known to those skilled in the art. The modem vendor can try at some specified data rate (or maximum data rate for rate adaptation) to substantially minimize the amount of power required for a given MAXSNRM. If the margin is less than MAXSNRM, but the generated and reported spectrum appears to be deviating, a controller such as a DSM center can select a PSD mask to monitor with bit swapping and bitloading in the present invention. Can be proposed.

Here, one processing procedure for loading will be described for use with embodiments of the present invention. Loading is essentially two vector quantities: a vector of incremental normalized energy whose elements are Δ (b) and a channel reference noise or channel normalized mean square error (MSE), vector of ν _n Depends on and. The latter quantity ν _n is

Where W _n is the frequency domain equalizer (FEQ) coefficient on the tone when FEQ is used. The energy of transmitting one or more bits on tone n when tone n is already transmitting bits of b _n is

However, the function Δ (b) does not depend on the tone index n, but depends on the ADSL group and the target margin TSNRM. The function Δ (b) defines the incremental (additional) energy required to transmit the b th bit on any channel with ν _n = 1 for the (b−1) th bit. . Therefore, the BCAP position is required (since it is not 15 or more in ADSL, Δ (16) = ∞ and Δ (0) = 0), by storing this function Δ (b), and the NSC channel dependence amount By calculating / updating and storing ν _n , n = 1,..., NSC, the extra energy for transmitting additional energy on either channel is reduced to 2 as in equation (3). It can be calculated by product of functions.

After calculating equation (3), the resulting total energy on the tone needs to be compared to any applicable MAXNOMPSD or PSDMASK limit, and this boundary is 2.5 dB (or , Smaller than 2.5 dB, the number of other accuracy bases that the designer can satisfy) or more, ΔE _n (b _n +1) = ∞ (ie, the incremental energy represents the loading performed by the processor) Can be reset to some maximum number). Such PSDMASK restrictions can often be imposed in ADSL systems. Such mask storage typically requires another 1 (just MAXNOMPSD) to about 20 locations (in ADSL2 +, PSDMASK levels for various tones). Table 1 in FIG. 9 shows that the G.G. When not used in 992.1 / 3/5 ADSL, the incremental energy Δ (b) and total energy when ν _n = 1 is not affected by PSDMASK and are listed (ADSL1 is , B = 1, so Δ (2) is the first interest for any tone n in ADSL1). The quantity ε is

Where CODEGAIN is 3.8 + 3 · (R / N) · b _ave dB when using only FEC, and an extra 6 parity bits in the gain Will be nominally about 3.8 dB, so that the loading is referred to for a system that has the total data rate as a reference, and this criterion has both parity and code. Will be treated as if it were not).

b _ave is the per-symbol that can be calculated from the total line rate estimate by dividing the total number of bits per symbol (BMAX) by the number of tones and then multiplying by the overhead percentage and nominal 3 dB / bit cost. Is the average (approximate) number of bits. Usually, this CODGAIN is about 5 dB. This extra gain greater than 3.8 dB is necessary because the line bit rate is loaded by the algorithm described here. In this case, the parity overhead amount (R / N) ≦. 08. The convention becomes optimistic if more parity is used and CODGAIN should not exceed 5 dB when loading the line bit rate. The 5 dB limitation can reduce the calculated line bit rate bit total BMAX (ie, the actual coding gain can be higher than 8 dB, but as high as the above equation represents). However, if a large parity part is used, the line is dominated by impulses or intermittent noise, and pessimistic coding gain is avoided.

  If the long-term line uses a large amount of parity, CODGAIN is usually high, but stationary noise does not dominate the performance, and accordingly, underestimating the coding gain is not a big error for the long-term line.

  In ADSL1, the tones are reordered for the progression of the encoder from the tone with the lowest number of bits to the tone with the highest number of bits. ADSL2 instead allows the receiver to reorder the tones for the progress of the transmitter encoder according to the receiver's desired and specified tone reordering. This tone re-ordering is communicated to the transmitter during training. In ADSL2 and ADSL2 +, reordering can be used to simplify the loading search algorithm, but embodiments of the present invention are implemented and used in a standardized manner where reordering is appropriate. Assume that it does not support accurate transmitter / receiver ordering. The reference loading algorithm described here is for all those for which ordering is specified.

  The following is an example of a reference training loading algorithm.

Two convergence criteria, namely
(1) the maximum final rate is realized, or
(2) Sequentially add each bit to the tone with the minimum ΔE _n (b _n +1) for all tones until one of the total energy allowed is satisfied.

  If decision condition (1) is met, all tones can increase energy to a ratio of PSDMASK (or MAXNOMPSD) applicable to each frequency. The smallest increase in such dB plus TSNRM is the actually reported SNRM. The minimum increase is such that the TSNRM plus increase exceeds MAXSNRM, so that all tones should instead increase energy by MAXSNRM-TSNRMdB.

  The following is an example of a standard SHOWTIME loading algorithm.

Find the tone with the minimum ΔE _n (b _n +1) for all tones at the current data rate and store the exponent n of the tone. Again, look for the tone with the maximum ΔE _m (b _m ) and save the index n. Swap bits from tone m to tone n if and only if ΔE _n (b _n +1) ≦ ΔE _m (b _m ).

  The total energy is saved if the margin does not exceed MAXSNRM for the new bit distribution. If MAXSNRM is now exceeded for this tone and this is the tone with the smallest margin, the energy for all tones should be reduced by the rate that the margin for tone n exceeds MAXSNRM.

  The ADSL1 and ADSL2 standards have a “bit swapping” mechanism that allows bits to move from tone m to tone n.

At the end of any processing procedure, the total energy for any tone is

Calculated by

The energy level, as will be correctly recognized by those skilled in the art, can be converted to the gain level g _n.

  Trellis coding imposes a slight complexity on loading, but follows the same basic principles. Loading for a DMT transmission system with trellis coding constitutes a group of subchannels each consisting of two tones. The two tones in a group are continuous tones, whatever ordering is used. In ADSL, when trellis coding is used, there will always be an even number of used tones, and thus a group of integers. The result is an incremental energy table that adds bits with trellis coding to the group of two tones (instead of adding to one tone). If BCAP = 15 for all tones, it can be loaded into one group up to 29 bits (15 for the first tone plus 15 for the second tone plus the extra required for trellis coding 1 bit).

Within the group of two tones, it can be assumed that ν _{n + 1} > ν _n without loss of generality (otherwise the exponent is re-established within the loading algorithm for computation). If you create and finish the loading algorithm, do not recreate the index). As a result, the incremental energy that adds a bit to a group is always the minimum energy that adds a bit in two tones, and starting with a group with 0 bits, the first information bit added is Actually, 2 bits added to tone n (the extra first 1 bit is for 1 bit / group overhead during trellis coding). Each subsequent additional bit takes one incremental energy unit instead of two incremental energy units for the first bit. The loading table is shown in Table 2 of FIG.

Thus, the loading algorithm in examining the group of tones (n n + 1) may, for each bit after the first bit added to tone n, possibly

Is used to add to individual tones.

To delete one bit, the 4D trellis loading algorithm instead

Check out.

If only trellis coding is used, CODGAIN in equation (4) should be 4.2 + 1.5 = 5.7 dB. If both trellis and FEC are used, CODGAIN should be 5.5 + 1.5 + 3 · (R / N) · b _ave , which can be estimated to be about 8 dB. . Again, (R / N) ≦. 08. The convention becomes optimistic if more parity is used and CODGAIN does not exceed 8 dB when loading the line bit rate. The 8 dB limitation can reduce the calculated line bit rate bit total BMAX, but if a large parity part is used, the line is dominated by impulse or intermittent noise, and the pessimistic coding gain is Be modest.

  Implementation of the above loading algorithm can result in “disjoint” or “sawtooth” energy characteristics due to “jumps” in energy having only an integer bit per tone. Thus, one reason for adjusting the gain is to “equalize” the margin for all tones. In general, this is a small effect, but it can produce a slightly higher modem / line margin, which does not need to be implemented by the modem manufacturer and often depends on the manufacturer. Not implemented. After the bit allocation is set, some tones may have a slightly higher margin than others (the reported margin is worst for all tones). In practice, the tone that will be added to the next bit and become one or more bits to load has the highest margin, the next to a different tone, the next highest margin, etc. The gains for those tones with higher margin than the last tone receiving one bit when loading will transmit slightly lower energy and more tones with the last loaded bit (unless the PSD mask is violated) can do.

Both ADSL1 and ADSL2 are designed so that the gain can be specified in the range [-14.5, + 2.5 + EXTGI] dB during training (where EXTGI = 0 is always for ADSL1). . ADSL2 allows the same range during SHOWTIME operation, accurate _{g n} values are transmitted over the overhead channels. ADSL1 has for the last established value or during the previous gain swap during training, -2 range during SHOWTIME, that limits the -1, 1 or 3dB change. Column gain swaps should not violate the range [-14.5, + 2.5 + EXTGI].

  The modem vendor should know that the ADSL1 modem has a synchronization symbol whose power is not reduced every 17 ms. That is, if the rest of the signal is reduced in power by some loading procedure (especially gain swapping, etc.), the intersymbol interference (ISI) from the ADSL1 synchronization signal is relatively large, May dominate all noise. This ISI can be canceled in a number of ways, including effectively reconfiguring the ISI and removing the ISI, since the synchronization symbols are known. Another solution is to change the time domain equalization setting as the gap between the sync symbol energy and the normal symbol energy increases. The ADSL1 standard does not mandate, but recommends that the SHOWTIME gain swap maintain all symbol energies at a constant symbol power spectral density level of +/− 2.5 dB.

  In the case of ADSL1, the basic loading step can follow the margin equalization step. This margin equalization step is performed after the data rate has been set (the desired rate has been achieved in a fixed rate ADSL or the maximum rate in rate adaptive ADSL has been determined).

The algorithm mentioned here can be used for ADSL1 margin equalization (ADSL1 gain swapping) during SHOWTIME, and for SHOWTIME gain swapping, the gain during training is such that the MAXSNRM does not exceed (MAXSNRM). If is exceeded during initial training, no gain swapping is required), assuming that it is already set. In the case of fixed rate loading, the energy used will be less than the total energy allowed (or the modem will retrain or possibly operate below TSNRM, in this case the procedure And should be used, resetting the TSNRM to the actual lower SNRM, the calculation of the margin for each tone is known to those skilled in the art.
1. Order tones from SN to MAX (and store order) with respect to SNRM [n] and store MAXS = maxSNRM [n] ≦ MAXSNRM.

  2. Each continuous tone is incremented by 1 dB to an index m, provided that SNR [m]> MAXS and MAXS → min (SNRM [m] +1, unless the total energy (or PSDMASK / MAXNOMPSD) is exceeded. MAXSNRM) Reset to dB. In the case of tones that will exceed PSDMASK / MAXNOMPSD, the tone is removed due to further problems.

  3. Again, as in step 1, reorder the tones (keep order).

  4). Repeat step 2 for the newly ordered tone.

  5. Repeat step 3.

  6). Repeat step 4.

  7). Undo all ordering and reinsert any tones that would have previously exceeded PSDMASK / MAXNOMPSD.

  Next, some gain swap is performed. By the end of this processing procedure, up to 3 dB is added to the tone having the minimum margin. The bit allocation does not change, but the margin can increase by 1, 2 or 3 dB for some tones. The new minimum margin will be more than before execution of the processing procedure, typically on the order of 1 dB.

  ADSL1 margin during initialization (and training or ADSL2 gain swapping for SHOWTIME), except that the same procedure as ADSL1 SHOWTIME can be used by designers in increments smaller than 1 dB, eg 0.5 dB Continues for equalization, so that the algorithm runs up to 5 passes (instead of 3), with each pass giving possibly another 0.5 dB for a subset of tones that become progressively smaller can do. Again, gain swapping is not required if MAXSNRM is exceeded (ADSL1 only) or achieved (ADSL1 or ADSL2).

  Of course, for all swapping methods, if the channel noise (MSE) varies with time, after performing the first bit swapping, the gain swapping can be performed again. Even if a bit swap is not required, a gain swap is required if the MSE changes.

The algorithm described above is easily modified for BCAP [n] and TSNRM [n]. For non-uniform BCAP [n] ≦ 15, Δ (b _n ) = ∞ for b _n > BCAP [n], which is obtained by multiplying Δ (b _n ) by ν _n and ΔE _n (b _n ) before getting. If trellis coding, BCAP [n], as in the case of Table 2 in FIG. 9, in the column of the last entry _{ν n ΔE n (BCAP [n} ]) = corresponds to ∞, [nu _{n + 1} of the column On the other hand, ΔE _{n + 1} (BCAP [n + 1] +1] = ∞ is applied to the information bits.

For non-uniform TSNRM [n]:

Therefore, it is simpler (less memory) to change each ν _n than to change Δ (b). However, TSNRM is a specified (usually 6 dB) single (uniform) margin. Both BCAP [n] and TSNRM [n] can be used for various improvements for a particular line.

  In the greedy algorithm, a tone that is already loaded into the maximum bit cap has a huge (large) cost to add additional bits, thereby avoiding bits / tones beyond the bit cap. The present invention recognizes that tones whose addition of bits exceeds the PSDMASK limit specified for them should also have enormous (large) cost at those locations, May or may not be implemented by a different manufacturer. Channel and / or noise changes are monitored and the above algorithm is continuously executed to allow the DMT transmission equipment to move the bits so that good energy usage and margins are maintained. The tremendous cost associated with exceeding the imposed spectral mask is that bits are allocated to the band limited by PSDMASK even if the band limited by PSDMASK becomes more attractive than other bands in theoretical water injection So that it is maintained during SHOWTIME operation.

PSDMASK, which inherently immensely costs additional bits on a particular tone when the current energy of that particular tone is already at or near the mask level, is used for bandwidth priority. In essence, due to the PSDMASK limitation, adding bits to the non-priority band is too costly and thus causes the discontinuous irrigation algorithm (ie, greedy algorithm) to load the same bits into the priority band. Energize or force, in this case, PSDMASK can be higher and, accordingly, can facilitate the loading of more bits. Accordingly, PSDMASK probably determines that the controller and / or dynamic spectrum manager (again, for example, an SMC or DSM center or manager) will find such bandwidth priority useful for the line. Therefore, it may be set well below the allowed MAXNOMPSD mask in some bands to indicate priority over the use of other bands. Centralized control of bit allocation is realistic because of the response speed required to change bit allocation when needed or appropriate to the effects of time-dependent channels (such as crosstalk changes). rare. Band priority, instead, perhaps specified in ADSL2 +, and the current, by deliberate choice of PSDMASK level proposed in VDSL2 (or by optionally alternatively tss _n level) the controller and Specified at initialization by the dynamic spectrum manager.

Energy E _n on a particular tone, the following

Where E _{0, n} is the REFPSD level. For example, an ADSL modem without power reduction and the use of PSDMASK would be known to both the transmitter and receiver—40 dBm / Hz (or as determined by various annexes, or implementation of the invention) It will have E _{0, n} corresponding to the other value specified by NOMPSD by the controller according to the configuration. The quantity g _n ² specifies the “g _i ” gain calculated by the receiver, which is typically −14.5 dB to +2.5 dB for the ADSL standard, and is theoretically not negative ( It can also be a (linear) value. The "gain" is the time of initialization exchange, or, either during "bit-swapping" in Modem dynamic Sakuchu, passed to the transmitter via a reverse control channel in DMT DSL. The tss _n ² parameter is fixed for any use of the modem by the MIB and can be 0-1. For water injection theoretical, tss, the gain parameter, any tss effects can not be performed, it is possible to set (if the gain g _n is not a limit) energy level to a desired injection level, most Useless. Of course, a value of tss = 0 avoids the use of tones and cannot be reversed. In practice, the upper limit of possible gain selection by 2.5 dB or by EXTGI allows tss to affect the limits of the loading algorithm and is therefore a useful tool. Specifically, the upper gain limit at +2.5 dB prevents significant reversal (if this limit is increased by EXTGI, more reversals are possible, and the value of EXTGI = 18.0 dB is probably wrong, May be used to re-introduce theoretical water injection band priority reversal.) Unmodified water treatment procedure recovers one band with a positive gain value when tss is low Will. The situation of E0 + 2.5 dB, the maximum value in equation (1) above, directly corresponds to the enormous (high finite accuracy) cost associated with adding additional bits to the tone in discontinuous loading. This non-linearity of discontinuous loading is important for bandwidth priority.

  In systems where the PSDMASK value is intentionally set low for other good bands, the enormous cost associated with loading more bits beyond the level that achieves masking in the band is Instead, it forces the loading algorithm to place bits in other available transmission bands that have not yet reached the PSDMASK PSD limit, giving priority to or prefer those other bands. This can also be done essentially using a non-standard BCAP [n] concept (variable bit cap with frequency). For example, in certain ADSL CO / RT mixing situations and embodiments upstream of VDSL, theoretical water injection looks more attractive (but limited by crosstalk generated by a second user on a longer line. Will be useless because it tries to keep loading in a lower frequency band. For the same situation, if the receiver also knows the PSDMASK setting, it should be set to a suitable low level (or bit cap is usually less than 15) to avoid crosstalk to longer lines. You will notice the enormous cost of exceeding PSDMASK (if kept in a small, constant number of maximum bits). Thus, discontinuous water injection will begin to add bits to the second higher frequency band on the shorter line in both embodiments and will have the same result as OSM. If the dynamic spectrum manager knows that a particular line is interfering (eg, using one of the estimation methods of the present invention described above), a significant improvement can be achieved. it can.

  In some embodiments of the present invention, PREFBAND is not only the cost of loading bits that are 1 or higher than PSDMASK (or effectively hindered), and MAXSNRM is not only the worst margin tone, but all Is a 1-bit unsigned integer that specifies what to apply to the tone. Those skilled in the art will appreciate that this embodiment of the present invention can also be used effectively for VDSL2. The determination of the PSDMASK level imposed is a matter of the controller such as the DSM center. The controller may need to know primarily the Hlog and noise power spectral density, as well as the Xlog of the DSM report that can be imposed on a bandwidth priority basis, primarily determining good PSDMASK levels. One method of such determination that avoids OSM advanced computation and convergence problems is all using Hlog, Xlog and noise power spectral density to find out which settings result in acceptable line performance improvements. Trying a few gross spectral levels and end-of-band frequencies with simulated iterative discontinuous irrigation for a number of channels.

  The OSM algorithm requires at least knowledge of crosstalk transfer capability between interfering lines for lower bandwidth priority. In ADSL2, or in any system (email / ftp) where all appropriate binder insertion loss and crosstalk insertion loss transfer functions are available to the controller or can be calculated by the controller, The central algorithm can determine the level to use for bandwidth priority. These levels are communicated to the ATU-C and / or ATU-R and are implemented by PSDMASK and / or bandwidth priority bits.

  Methods such as forward error correction and bit error rates in DSL systems are intended to help ensure that DSL services provide high data rates to DSL customers. (In some cases, defined as a small number of retraining, reduced opportunities for reduced throughput due to high code violation (CV) count, etc.) “Reliability” and “High data rate” are both important, but both There has never been a clear way to communicate, and service providers often have a mix of unreliable lines (ie, frequency retraining, high CV counts) and unachieved lines (low data rates). Become. Operators also find that their reliability and service / performance issues contribute to customer satisfaction and loyalty, including repair needs and costs (eg, truck rolls), and customer transfer rates. ing. Embodiments of the present invention provide a method and technique for achieving a desired data rate with a minimum margin, or otherwise maximizing a desired data rate and maintaining one or more minimum reliability situations. including. That is, adaptive power / margin control can be used to optimize performance, using embodiments of the present invention that can maintain the maximum data rate (for good lines). Adaptive data rate control adapts to one or more performance-related parameters and / or targets (eg, CV count, retraining, etc.) if the maximum data rate can be achieved (for insufficient lines) It can be implemented to achieve the best possible data rate.

  In current systems and standards, the data rate is selected by the modem during training (or even during SHOWTIME in ADSL2) when dynamic rate adaptation is “on”. This selected data rate cannot exceed the maximum rate for which the customer pays for it and is therefore fixed at this maximum rate on a good line. For insufficient lines, the rate is less than the maximum rate and is selected for a predetermined margin level (eg, 6 dB) during training. In other words, adaptive data rate control is applied only when the line cannot achieve the maximum rate (eg, 1.5-3 Mbps) within the customer's service plan.

  If the maximum data rate can be reliably achieved with an appropriate margin, the data rate does not need to be adaptively controlled (the maximum rate is simply selected); instead, a margin is provided to avoid excessive levels. Control (and maintaining the maximum rate) becomes a problem. In such a case, the margin target for power reduction may allow the desired performance parameter and / or target (eg, retraining / CV criteria) to accommodate the line and minimize interference to other lines. In addition, it may be necessary to determine adaptively for each line. For example, for a line that does not experience any fluctuations in noise power, a 6 dB margin target may be sufficient. However, for lines that experience noise power fluctuations up to 10 dB, the appropriate margin to select may be 16 dB. However, until now, fixed margin targets have been used during training, and only noise power during training has been used as a criterion for selecting power reduction (ie, noise power history). Or the distribution was not considered).

  In the current implementation, similar problems exist for insufficient lines. If the maximum data rate cannot be reliably achieved with an appropriate margin, the power need not be adaptively controlled (the maximum power is simply selected). Therefore, controlling the data rate (and maintaining the margin level) to avoid an excessive rate becomes a problem instead. For example, for a line that does not experience any fluctuations in noise power, a data rate of 2.0 Mbps may be appropriate to accommodate the desired retraining / CV or other performance related criteria. However, for lines that frequently experience greater noise power, a data rate of 1.6 Mbps may be appropriate instead. However, to date, a fixed margin target has been used during training, and only the noise power during training has been entered to select the data rate (ie the history or distribution of noise power is Was not considered,). In the following examples of embodiments of the present invention, methods and techniques that utilize history and / or distribution are described.

For the ADSL line in question, various operational data can be collected regularly or irregularly. This data may include current margin, current data rate, current maximum achievable data rate, FEC correction count, CV count, retraining count, channel transfer function and noise spectrum. Also, using the collected operational data, marginal probability distribution, maximum achievable data rate, FEC correction count, CV count, retraining count, etc. can be estimated as a function of data rate. The channel transfer function and noise spectrum may not be immediately available to the controller if operational data is collected only from the ATU-C side of the ADSL system, but at least some of the useful data Can be estimated in such situations. Such methods of obtaining estimates, Ru can be found in the files US Patent Application No. 10 / 817,128 Pat April 2, 2004.

One example of a margin distribution plot is shown in FIG. For a particular collected data rate R _COLLECT , eg, 3 Mbps, operational data is collected over time to determine the percentage of time that the DSL line is operating at R _COLLECT during that time. In the example of FIG. 10, the DSL line operates at 3 Mbps for about 50% of the time with a 16 dB margin. Similarly, for the same collected data rate, a DSL line uses a 10 dB margin for about 10% of the time and a 4 dB margin for about 1-2% of the time. By adding the total ratio with respect to the predetermined margin range, it is possible to determine the possibility of operation at the maximum margin at the rate or below the maximum margin.

Margins are closely related to CV count, retraining rate, maximum data rate and other related performance parameters. For example, a high CV count and / or retraining rate may be statistically related to the required track roll for a given DSL line. Similarly, customer satisfaction (eg, as measured by the number of customers that stop a given operator's service) may also be statistically related to CV count and / or retraining rate, as well. is there. Therefore, after finding one or more distributions of performance parameters as a function of the data rate, the probability that the count of probability and CV of line failure as a function of the data rate (line retraining of) exceeds a threshold value calculated can do. If performance thresholds / targets are of particular importance to an operator or other party, the present invention allows the party to adaptively control data rate and / or power / margin usage and Be able to handle one or more of the targets.

  As a result, the maximum power drop (minimum margin) or maximum data rate can be selected and reliability criteria (eg, number of retrains and CV count exceeding specified threshold) can be met.

  For example, multiple thresholds can be used for retraining number and CV count, and the criteria can be as follows:

-Number of retraining (per day) <1 with a probability of 50% or more,
-Number of retraining (per day) <3 with a probability of 90% or more,
-Number of retraining (per day) <99% with a probability of more than 99%,
-CV count (with 15 minute period) <99% with a probability of 2000,
-CV count (with 15 minute period) <95% with a probability of 1000,
-500 with a probability of CV <90% (with a 15 minute period).

  The highest data rate that meets the maximum power drop or all six criteria is selected.

A method 1100 according to one embodiment of the invention is shown in FIG. Initially, at 1110, operational data is collected for one or more data rates R _COLLECT . At 1120, using this collected data, one or more distributions of performance parameters (such as margins as shown in FIG. 10) are plotted as a function of a predetermined data rate used to collect the data. The The highest data rate R that meets one or more performance targets is then selected at 1130. If at decision 1140, the highest data rate that meets the performance target is the maximum data rate (R _MAX ), the performance parameter is optimized at 1150 to maintain the maximum rate (eg, reduce margin or Strengthen power reduction). If the highest rate that meets the performance target at 1130 is not R _MAX , the DSL line operates at the selected R, and the performance parameters follow its distribution. In order to ensure that confidence in one or more distributions is valid, the system can be updated by itself, as shown in FIG.

  As reflected in FIG. 10, there is an overall trade-off between the margin and the data rate, and the value of the margin decreases as the data rate increases. Using the estimated and / or collected information, the controller is forced for a given margin value based on the same data set used to estimate the distribution of various performance parameters for a given data rate. DSL direct performance parameter distribution such as retraining, CV, code error correction, etc. can be found. The retraining and CV criteria can then be interpreted or explained in terms of margins. For example, the following margin criteria can be applied to a line using retraining and CV criteria.

  -The margin needs to be 3 dB or more for 99% of the time.

  -The margin needs to be 5 dB or more for 95% of the time.

  -The margin needs to be 6 dB or more for 90% of the time.

  Based on the margin distribution, the maximum power reduction or maximum data rate can be selected such that all three criteria are met for the DSL line. It is also possible to combine the six retraining and CV criteria with the three margin criteria so as to reduce the risk associated with estimation errors.

  Also, embodiments of the present invention are equally applicable to situations where a line experiences two different line conditions with long stay periods in both two different states. In such a case, two sets of margin distribution criteria can be configured, and an appropriate set of criteria can be used when detecting the current state. Obviously, the present invention can be extended to lines with more than two states.

  In general, embodiments of the present invention employ various processes that require data stored in or transferred via one or more modems and / or computer systems. Embodiments of the invention also relate to hardware devices or other devices that perform these operations. This apparatus may be specially configured for the required purposes, or it may be a general purpose computer selectively activated or reconfigured by a computer program and / or data structure stored in the computer. The processes presented herein are not inherently related to any particular computer or other apparatus. In particular, various general purpose machines can be used with programs written in accordance with the teachings herein, or more advantageous to construct a more specialized apparatus that performs the required method steps. There is a possibility. Specific structures for a variety of these machines, based on the description given below, will be clearly understood by those skilled in the art.

  Embodiments of the present invention as described above employ various process steps that require data stored in a computer system. These steps are those requiring physical manipulation of physical quantities. In general, though not necessarily, these quantities take the form of electrical or magnetic signals that can be stored, transferred, combined, compared, and otherwise processed. It is convenient to refer to these signals as bits, bitstreams, data signals, command signals, values, elements, variables, characters, data structures, etc. mainly for general use. However, it should be remembered that all of these and similar terms are to be associated with the appropriate physical quantities and are simply convenient labels used for those quantities.

  Further, the operations performed may be explicitly referred to as identifying, adjusting or comparing. In any of the operations described herein that form part of the present invention, these operations are machine operations. Useful machines for performing the operations of embodiments of the present invention include general purpose digital computers, processors, modems or other similar devices. In all cases, a distinction should be made between the method of operation when operating a computer and the method of computation itself. Embodiments of the invention relate to method steps for operating a computer in processing electrical or other physical signals to generate other desired physical signals.

  The embodiments of the present invention further relate to a computer-readable medium including program instructions for executing various operations executed by a computer. The media and program instructions may be specially designed and configured for the present invention, or they may be known ones available to those having skill in the computer software arts. Good. Examples of computer-readable media include, but are not limited to, magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs, magneto-optical media such as floppy disks, and read-only storage elements. Hardware devices specially configured to store and execute program instructions, such as (ROM) and random access memory (RAM). Examples of program instructions include, for example, both machine code generated by a compiler and files containing high level code that can be executed by a computer using an interpreter.

  FIG. 8 illustrates an exemplary computer system that can be used by a user and / or controller in accordance with one or more embodiments of the present invention. Computer system 800 is coupled to a storage device that includes a primary storage device 806 (typically random access memory, or RAM), and a primary storage device 804 (typically a read-only storage element, or ROM). It includes a number of processors 802 (also called a central processing unit or CPU). As is known in the art, primary storage 804 operates to transfer data and instructions in one direction to the CPU, and primary storage 806 typically transfers data and instructions bidirectionally. Used to do. Both of these primary storage devices may include any suitable computer readable media as described above. Mass storage device 808 is also coupled bi-directionally to CPU 802, provides additional data storage capacity, and may include some of the computer-readable media described above. The mass storage device 808 can be used to store programs, data, and the like, and is typically a secondary storage device such as a hard disk that is slower than the primary storage device. It will be appreciated that the information stored in the mass storage device 808 may be incorporated into a portion of the primary storage device 806 as a virtual storage, if appropriate, in a general manner. Certain mass storage devices such as CD-ROMs may also flow data to the CPU in one direction.

  The CPU 802 can also be a video monitor, trackball, mouse, keyboard, microphone, touch display, transducer card reader, magnetic or paper tape reader, tablet, stylus, voice or handwriting recognition device, or, of course, other computers, other computers, etc. Coupled to an interface 810 that includes one or more input / output devices, such as known input devices. Ultimately, CPU 802 may couple to a computer or telecommunications network, depending on the circumstances, using a network connection, generally as indicated at 812. Connection 812 can be used to communicate with the DSL system and / or the modem. In some cases, the computer system 800 is proprietary to the DSL system, perhaps through operator equipment (eg, CO) or in any other suitable manner (eg, connection of a given DSL system to the NMS). May have dedicated and / or other specific connections. For such a connection, the CPU can receive information from the network and / or DSL system, or output information to the network and / or DSL system in the course of performing the method steps described above. It is intended to be able to. The above-described devices and components will be familiar to those of skill in the computer hardware and software arts. The hardware elements described above may form a number of software modules that perform the operations of the present invention. For example, the instructions for executing the margin monitoring and control controller can be stored on the mass storage device 808 (which may be a CD-ROM or may include a CD-ROM), primary The program is executed on the CPU 802 in conjunction with an appropriate computer program product used in the storage device 806 and the system 800. In the preferred embodiment, the controller is divided into software sub-modules.

  Many features and advantages of the invention will be apparent from the description, and the appended claims are intended to cover all such features and advantages of the invention. Further, since many modifications and variations will occur to those skilled in the art, the present invention is not limited to the exact construction and operation as shown and described. Accordingly, the described embodiments are not limiting and should be taken as illustrative, and the present invention should not be limited to the details set forth herein, but is predictable now or in the future. Whether or not should be defined by the following claims and their full scope of equivalents.

This is a general block reference model system compliant with the G.997.1 standard. FIG. 2 is a schematic block diagram illustrating a general and exemplary DSL arrangement. It is a schematic block diagram of one embodiment of the present invention in a DSL system using a controller such as a DSM center. FIG. 3 is a comparison diagram of power adaptation implementation of a DSL system. FIG. 6 is a comparison diagram of rate adaptation implementation of a DSL system. It is a comparison figure of margin adaptation implementation of a DSL system. FIG. 2 is a flow diagram and schematic diagram illustrating the operation of an ADSL1 system according to an embodiment of the present invention. FIG. 2 is a flow diagram and schematic diagram illustrating the operation of an ADSL1 system according to an embodiment of the present invention. FIG. 2 is a flow diagram and schematic diagram illustrating the operation of an ADSL2 system according to an embodiment of the present invention. FIG. 2 is a flow diagram and schematic diagram illustrating the operation of an ADSL2 system according to an embodiment of the present invention. FIG. 4 is a flow diagram illustrating a method according to an embodiment of the invention. 1 is a block diagram of an exemplary computer system suitable for implementing embodiments of the present invention. A set of bit loading energy tables. FIG. 5 is an example of a margin distribution for a predetermined data rate estimated based on collected operational data. FIG. 2 is a method according to an embodiment of the invention using an estimated distribution of one or more performance-related parameters such as margins. 1 is an embodiment of the present invention showing a “high performance” modem unit having a controller with a processor and memory integrated with a DSL modem.

Claims (29)

A method for controlling a margin related parameter set in a DSL (Digital Subscriber Line) system comprising:
And that modem pair is be to collect operational data from the DSL system operating, the operating data with historical data relating to past performance and compliance past margin of the modem pair,
Analyzing the collected operational data including the historical data ;
Generating the margin related parameter set comprising at least one margin related parameter value used by the DSL system , the at least one margin related parameter value assisting the modem pair to meet a target margin; To be calculated as
Instructing at least one modem of the modem pair of the DSL system to operate using the margin related parameter set having the at least one margin related parameter value ;
A method comprising: The operating data is data rate data, SNR margin data, the maximum achievable data rate, total transmit power data, the code violation count data, forward error correction data, error data, retraining number data or close said DSL system, The method of claim 1, comprising at least one of data relating to crosstalk between DSL lines. Analyzing the operation data, current SNR margin, maximum SNR margin, past SNR margin, minimum SNR margin, the current data rate, maximum data rate, historical data rate, minimum data rate, the current total transmit power, the maximum Aggregate transmitter power , past aggregate transmit power , past number of code violations , code violation threshold, past forward error correction number , forward error correction threshold, past retraining number , and retraining threshold The method of claim 1, comprising analyzing at least one of the values . Wherein said at least one margin-related parameter value is used by the DSL system, the maximum power spectral density level, the maximum transmission power level, the target SNR margin, the maximum SNR margin, the minimum data rate, said maximum data rate, and MAX At least one of the preferred band (or “PREFBAND”) bits or indications sent to the modem pair of the DSL system to indicate that the SNR M parameter should be applied to the margin detected on all tones 4. The method of claim 3, comprising one. The generated margin-related parameter set includes Maximum Nominal Power Spectral Density or MAXNOMPSD, Maximum Nominal Aggregate Transmitted Power or MAXNOMATP, gain or gi, bit loading (bit -loading) or bi, Transmit Spectrum Shaping or tssi, calculated power drop or allowed power cutback value or PCB, Maximum Received Power or MAXRXPWR, power Power spectral density or PSDMASK, carrier mask or CARMASK, preferred band or PREFBAND, Target Signal-to-Noise Ratio Margin or TARS NRM, Minimum Signal-to-Noise Ratio Margin or MINSNRM, Maximum Signal-to-Noise Ratio Margin or MAXSNRM, Bit-CAP or BCAP [ n] (n represents a tone) , Target Signal-to-Noise Ratio Margin or TSNRM [n] (n represents a tone) , maximum transmit power spectral density level, allowable gain value, bit allocation, power spectral density mask, the data relating to crosstalk between DSL lines and the adjacent modem pair of DSL systems, transmission bandwidth priority, bit cap for each tone, the target SNR margin per tone, the minimum data rate or the maximum data The method of claim 1, comprising at least one of the rates. Analyzing the collected operational data;
Generating a distribution of at least one performance-related parameter included in the collected operational data, and calculating at least one probability from the generated distribution;
The method of claim 1, comprising: The performance-related parameters include a past SNR margin, a past data rate, a past maximum achievable data rate, a past number of code violations, a past forward error correction count, a past error data count, or a past retraining count. The method of claim 6, comprising at least one of: Generating the margin related parameter set includes
Comparing the at least one calculated probability to a threshold ;
Updating at least one margin related parameter value in the margin related parameter set based on the comparison of the at least one calculated probability to the threshold ;
The method of claim 6 comprising: Said comparing at least one of the computed probability threshold, comparison to the margin threshold probability calculated by the margin, compared for the code violation threshold for probability computed code violations, the 9. The method of claim 8, comprising at least one of them. The DSL system, D SL system training before, during DSL system training, prior to the second training state after the first DSL system training state, during normal DSL system operation or "SHOWTIME", or, normal The method of claim 1, wherein the method is instructed to use the margin related parameter set at one or more of periodically during DSL system operation. A method for controlling a margin related parameter set of a first DSL (digital subscriber line) system, the margin related parameter set comprising at least one margin related parameter value, the method comprising:
Collecting DSL operational data from the first DSL system in which a modem pair operates, the operational data including historical data related to past performance and past margin compliance of the modem pair ;
Analyzing the collected DSL operational data including the historical data ;
Generating the margin related parameter set having at least one margin related parameter value calculated to assist the modem pair in meeting a target margin ; and
Instructing at least one modem of the modem pair of the first DSL system to operate using the generated margin related parameter set having the at least one margin related parameter value . Collecting DSL operational data and analyzing the collected operational data is a dynamic spectrum manager or other controller that manages the DSL system is a service provider that is completely independent of the system operator of the DSL system or 12. The method of claim 11, comprising managing the DSL system as an operator . The operating data is data rate data, SNR margin data, the maximum achievable data rate data, total transmission power data, the code violation count data, forward error correction data, error data, retraining number data, channel attenuation data, noise 12. The method of claim 11, comprising at least one of power spectral density data and data relating to crosstalk between a modem pair of the DSL system and an adjacent DSL line . The method of claim 11, wherein analyzing the collected operational data comprises evaluating data related to crosstalk between a modem pair of the DSL system and an adjacent DSL line .   Analyzing the operational data collected from the first DSL system and the second DSL system is an iterative water-filling simulation for the first DSL system and the second DSL system. 13. The method of claim 12, wherein the input to the simulation comprises Hlog, Xlog, and noise power spectral density. Margin The margin-related parameter set, the maximum power spectral density level, the maximum transmission power level, the target SNR margin, the maximum SNR margin, the minimum data rate, maximum data rate, and the MAXSNRM parameter is detected in all tones 15. The method of claim 14, comprising at least one of a suitable band (or "PREFBAND") bit or indication transmitted to a modem pair of the DSL system to notify that it should be applied to . The generated margin-related parameter set includes Maximum Nominal Power Spectral Density or MAXNOMPSD, Maximum Nominal Aggregate Transmitted Power or MAXNOMATP, gain or gi, bit loading ( bit-loading) or bi, Transmit Spectrum Shaping or tssi, calculated power drop or allowed power cutback value or PCB, Maximum Received Power or MAXRXPWR, Power spectral density or PSDMASK, carrier mask or CARMASK, preferred band or PREFBAND, Target Signal-to-Noise Ratio Margin or TAR SNRM, Minimum Signal-to-Noise Ratio Margin or MINSNRM, Maximum Signal-to-Noise Ratio Margin or MAXSNRM, Bit-CAP or BCAP [ n] (n represents a tone) , Target Signal-to-Noise Ratio Margin or TSNRM [n] (n represents a tone) , maximum transmit power spectral density level, allowable gain value, bit allocation, power spectral density mask, data relating to crosstalk between DSL lines adjacent to modem pair of the DSL system, transmission bandwidth priority, bit cap for each tone, the target SNR margin per tone, the minimum data rate or maximum, The method of claim 11, comprising at least one of data rates. The first DSL system, prior to the D SL system training, during the DSL system training, prior to the second training state after the first DSL system training state, during normal DSL system operation or "SHOWTIME" 12. The method of claim 11, wherein the method is instructed to use the margin related parameter set in one or more of, periodically, during normal DSL system operation. A controller for a DSL set comprising one or more DSL (Digital Subscriber Line) systems,
A collection module configured to collect operational data from at least one DSL system in which a modem pair operates, wherein the operational data includes historical data related to past performance and past margin compliance of the modem pair. When,
An analysis module coupled to the collection module and configured to analyze the collected operational data including the historical data ;
A margin related parameter set coupled to the analysis module and having at least one margin related parameter value calculated to assist the modem pair in meeting a target margin is transferred to at least one DSL system in the DSL set. A command signal generation module configured to apply, wherein the margin related parameter set is based on the analysis of the collected operational data including the historical data ;
Controller with. The collection module, modem before Symbol DSL set, the DSL set management entity, the DSL set of MIB, network management system coupled to the DSL set, broadband network are coupled to the DSL set or the DSL set, The controller of claim 19, wherein the controller is coupled to at least one of the databases storing the operational data of It said analysis module, before Symbol collection module or database for storing operation data of the DSL set, is coupled to at least one of the controller of claim 19. The command signal generation module, before Symbol DSL set of modems, the DSL set management entity, the DSL set of MIB, the network management system coupled to the DSL set or a broadband network that is coupled to the DSL set, the 20. The controller of claim 19, coupled to at least one of them. The controller of claim 19, wherein the DSL set comprises at least one of an A DSL1 system, an ADSL2 system, or an ADSL2 + system, or a VDSL system. The command signal generating module, before SL prior training in DSL system within DSL set during training in DSL systems in the DSL set in, the first training state at DSL system within the DSL set Later before the second training state in the DSL system in the DSL set, during normal operation of the DSL system in the DSL set or “SHOWTIME” , or normal in the DSL system in the DSL set 20. The controller of claim 19, wherein the controller is configured to generate a command signal at one or more of periodically during operation. Said controller, at least one of the at least directly attached processing unit to one of the modems, DSL optimizer, DSM Center or a dynamic spectrum manager,, in D SL system, network management systems, the DSL set The controller of claim 19, which is a component in The operation data, data rate data, SNR margin data, the maximum achievable data rate data, total transmission power data, the code violation count data, forward error correction data, and error data, retraining number data, and the first DSL system 20. The controller of claim 19, comprising at least one of data relating to crosstalk between one or more adjacent DSL lines. A calculation configured to calculate at least one distribution of operating data, calculate at least one probability associated with the operating data, and compare the at least one calculated probability with a threshold value ; The controller of claim 19 comprising a module. The controller of claim 19, wherein the analysis module comprises a calculation module configured to perform a comparison between the collected operational data and one or more thresholds . The margin-related parameter set includes Maximum Nominal Power Spectral Density or MAXNOMPSD, Maximum Nominal Aggregate Transmitted Power or MAXNOMATP, gain or gi, bit-loading ) Or bi, Transmit Spectrum Shaping or tssi, calculated power drop or allowed power cutback value or PCB, Maximum Received Power or MAXRXPWR, power spectrum density (Power spectral density) or PSDMASK, carrier mask or CARMASK, preferred band or PREFBAND, Target Signal-to-Noise Ratio Margin or TARSNRM, Minimum Signal-to-Noise Ratio Margin or MINSNRM, Maximum Signal-to-Noise Ratio Margin or MAXSNRM, Bit-Cap or BCAP [n] (N represents a tone) , Target Signal-to-Noise Ratio Margin or TSNRM [n] (n represents a tone) , maximum transmission power spectral density level, allowable gain value, bit allocation the power spectral density mask, the data relating to crosstalk between DSL lines and the adjacent modem pair of DSL systems, transmission bandwidth priority, bit cap for each tone, the target SNR margin per tone, the minimum data rate, maximum data rate, Or the SNR margin measured on every used tone is the maximum SNR margin value The controller of claim 19, comprising at least one of a PREFBAND indication that can be required to not exceed.

Images