TIE 1.4/97-362
`
`TIE1.4
`Meeting Report
`September 22-25,1997
`Minneapolis, Minnesota
`
`Chair:
`Vice Chair
`Secretary:
`
`Tom Starr (Ameritech)
`Massimo Sorbara (G/obeSpan)
`Ken Hohhof (Westel/)
`
`1. OVERVIEW
`
`158 persons attended this TlE1.4 meeting, which was held September 22-25, 1997.
`
`The Working Group considered:
`
`4 Spectral Compatibility contributions
`
`19 HDSL2 contributions
`
`55 ADSL contributions (Plus 2 deferred to the December meeting)
`
`16 VDSL contributions (Plus 12 deferred to the December meeting)
`
`1 Breakout Group report (on ADSL Initialization and Timing)
`There were no contributions on HSDL Iss. 2 Technical Report, ISDN Basic Rate U or srr Interfaces, Baseband
`Digital Data At 64 Kb/s and Below (DDS), V-Interface, or CSDC Network Interface.
`
`2. BUSINESS
`
`2.1 CALL TO ORDER
`
`The meeting was called to order by the Chair at 8:38 AM on Monday, September 22, 1997. The host, ADC, was
`thanked for providing facilities for the meeting.
`
`2.2 ANTITRUST NOTICE
`
`The Chair reviewed Committee Tl antitrust policies. WG members were asked to review their contributions and
`advise the Chair of any material that might be in violation of antitrust guidelines.
`
`2.3 DISTRIBUTION OF CONTRIBUTIONS
`
`The contributions were distributed according to contribution list TlE1.4/97-OOORl. Several additional
`contributions were added to the list at this time and distributed.
`
`3. LIAISONSIMEETING REPORTS
`
`TlAI Dick Bobilin, Creative Communications, reported that TlAI met August 4-8 in Kansas City. TlA1.2
`resolved all letter ballot comments on the draft Supplement to TR24 (Network Survivability Perfonnance)
`and the draft revision of Tl.511 (B-ISDN A TM Layer Cell Transfer Perfonnance). TIAl.3 resolved all
`letter ballot comments on the draft revision ofT1.511 (B-ISDN ATM Layer Cell Transfer Perfonnance).
`These 3 documents are now approved by Committee Tl and will be sent to ANSI for publication. TlAl.3
`made good progress on the development of a draft standard on "Internet Protocol (IP) Data Communication
`Service - User Information Transfer and Availability Perfonnance Parameters". TlA1.5 agreed to send out
`a letter ballot withdrawing Tl.314 (Digital Processing of Video Signals - Video Coder/Decoder for
`
`1
`
`

`
`"
`
`TIE 1.4/97-362
`
`Audiovisual Services at 56 to 1536 kbitls) which is being replaced by ITU-T H.26 1. Letter ballots for new
`TIA1.7 projects on "Objective Measures for the Assessment of Audio Quality" and "Interaction between the
`psrn and other networks and terminals" both passed. The next meeting ofTIAI will be October 27-31 in
`Austin, Texas.
`
`TIMI Dick Bobilin, Creative Communications, reported that TIMI met August 11-15 in Torrance, California.
`TIM1.3 resolved all comments from the second default letter ballot on the revision of T1.231 (Layer I In(cid:173)
`Service Digital Transmission Performance Monitoring), which is now approved by Committee TI and will
`be sent to ANSI for publication. It was agreed that T1.217 (ISDN Management - Prinuuy Rate Physical
`Layer) will be sent out for reaffirmation. The next TIMI meeting will be November 3-7 in Providence,
`Rhode Island.
`
`TISI Dick Bobilin, Creative Communications, reported that TISI met August 4-8 in Kansas City. TISI
`resolved all letter ballot comments on the draft Intelligent Network standard, and a default TI letter ballot
`will be sent out as soon as the revised text is available. The next meeting of TIS I will be November 17-21
`in Dallas, Texas.
`
`TIXI Dick Bobilin, Creative Communications, reported that TIXI met June 24-27 in Minneapolis. TIX1.3
`developed new technical input to the Issue 3 text for Tl.lOl (Synchronization Interfaces). It was agreed
`that Stratum 3 clock specifications will be added to T 1.1 0 I. The entry into holdover transient was discussed
`and modifications were made to the TI.I 0 I annexes. A living list for synchronization status messaging is
`being developed which identifies topics to be studied. It was noted that TR33 (Synchronization Status
`Messages) needs to be reopened. Work is beginning on the development of two new Technical Reports on
`"Effect of ATM Network Timing on Constant Bit Rate Services" and "Synchronization Network
`Architecture". TIX1.5 dealt with patent issues concerning the draft Issue 3 Tl.105.01 SONET Automatic
`Protection Switching standard, which was approved for TI balloting. The development of a new
`Supplement to Tl.105.07 (SONET - Sub STS-I Interface Rates and Formats Specification) which includes
`nxVf Group interfaces was completed and approved for TIletter balloting. The next meeting ofTIXl will
`be October 13-17 in Naperville, Illinois.
`
`TIAG Ed Eckert talked about electronic document distribution. It has been made a priority to develop a system
`which allows document number assignment and uploading via web browser forms, without dealing with
`things like .tid files. In response to a question about viruses, he noted that a nightly scan has always been
`done on the server, but updating of the virus definition files had lapsed. This has been corrected, and the
`VDFs will be updated monthly. Also, virus scanning has been added to the upload directoty - files will be
`scanned before being posted to the working group directories.
`--- - - -
`
`TR41 No report at this meeting.
`
`ITU-T Dick Stuart, 3COM, reported that Q4/SGI5 met'in July in Columbia, Maryland. The group agreed to start
`work on series of recommendations:
`
`• G.DMT (fashioned after T.413 Iss. 2,)
`• G.Lite
`• G.HS (common handshaking recommendations for xDSL)
`• G.Test
`
`G.CAP and G.QAM were also considered, but no work was started.
`
`There was a proposal to have interested parties meet some evening during the week, but it was decided that
`a meeting restricted to ITU members would conflict with the ANSI open meetings policy. Les Brown asked
`whether TIEl.4 should add ADSL Lite to its agenda, and suggested that if we intend to have input to the
`ITU terms of reference, we should not delay. Tom Starr clarified that the term ADSL Lite is used to
`describe 2 concepts - reduced complexity ADSL, and simplified CPE environment by removing POTS
`splitters. John Bingham noted that the main intent of G.Lite was to simplify the CPE end, while retaining
`
`2
`
`

`
`TIE1.4/97-362
`
`compatibility with standard ADSL at the CO end. Gordon Bremer noted that ITU hasn't determined the
`implementation, which could be HDSL Lite, but Dick Stuart said all discussion to date has been on ADSL.
`Dick also noted that there may be a trademark close to ADSL Lite, and he suggested use of the term G.Lite.
`Tom Starr suggested that the WG take an active role and coordinate with ITU. There were no objections,
`and the WG agreed to consider both meanings of ADSL Lite for Iss. 3, but not to discuss ADSL Lite at this
`meeting which concentrated on Iss. 2. Chris Hansen agreed to work with others to draft a liaison to ITU
`saying TIE1.4 will work on ADSL Lite and desires close cooperation.
`
`The next meeting ofQ4/SGI5 will be October 27-29 in Red Bank, New Jersey. Note that only ITU-T
`members may attend or participate on conference calls. Work is posted on a reflector and ftp site.
`
`ETSI Hans-Jorg Frizlen, Ericsson, reported that ETSI TM6 met in Berlin, Germany, June 2-6.
`
`TM6 dealt with all their four work-items at this meeting. For all this work tight cooperation with TlE1.4 is
`desirable.
`
`Basic Rate ISDN: A Living List captures the items that were unresolved with the last revision ofETR80 to
`stabilize these points and to give visibility to future potential revision points. It is the preference of TM6
`that TIEl, as originator of the ISDN-BA
`
`HDSL: A proposed single pair HDSL system with adaptive bitrates was discussed again, but due to
`insufficient support no work-item was created. The proposal shall be kept under study. TM6 intends at next
`weeks meeting to edit and reformat ETR152 into an ETS.
`
`ADSL: In principle the ETSI specification on ADSL is similar to T1.413 version 1 with some European
`additions as 2 Mb/s and different test loops. Even the version 2 of T1.413 may be treated in a similar way.
`Additional to this TM6 has two study work items dealing with ADSL and ISDN BA on the same pair, and
`the study of higher bitrates for future systems
`
`TIE1.4/97-305 (Liaison from ETSI WG TM6)
`
`Discussion: ADSL startup currently requires around 10 seconds, and while this is acceptable at service
`turn-up, some operators are concerned about losing service for this long after a "mini-interruption", and
`would like a warm-start procedure that takes around 100-300 ms. TM6 prefers that this be captured in
`T1.413 Iss. 2, but realistically realizes that it may fall into Iss. 3.
`
`Resolution: This is already on the Iss. 3 living list (item I)-a&Vnder Study.
`
`VDSL: The biggest work in TM6 is momentarily on the VDSL system. A specification covering functional
`requirements, network characteristics, service needs, bitrates, RFI will be finalized to the end of this year
`(an ETS can be approved more quickly than an ETR). Following this, electrical characteristics including
`line code will be developed.
`
`Future meetings:
`
`September 29 - October 3
`November 17-21
`January 26-30, 1998
`April 20-24
`June 22-26
`September 21-25
`
`Lannion, France
`Verona, Italy
`tbd (probably Spain)
`tbd (probably Belgium)
`tbd (probably Sweden)
`tbd
`
`Hans-Jorg announced his retirement from Ericsson at the end of the year as he will reach 65. There is some
`possibility that he may continue as chair for ETSI TM6 iffunding is allocated to this work. Tom Starr
`recognized Hans' long service as liaison to TlE1.4 and its predecessor, TlE1.3.
`
`3
`
`

`
`TIE 1.4/97-362
`
`IEEE P.1007
`
`No report at this meeting.
`
`IEEE 802.14
`
`No report at this meeting.
`
`ADSL Forum Gavin Young, BT Labs, gave an informal report. ADSL Forum met the previous week in
`Brussels. There were previously 7 technical groups, which have been reduced to 5. The network migration
`group finished its Working Text, which was sent out for letter ballot. The SNAG (architecture) group
`merged into the packet mode and A TM groups. The CPE group will be renamed because it also covers the
`CO splitter. The test group was renamed operations and test. The CPE and packet groups are putting
`Working Texts out for ballot, and the network management group will be publishing a MIB soon.
`
`ATM Forum
`
`ATM Forum was meeting the same week. No report at this meeting.
`
`DA VIC
`
`DA VIC met the previous week. No report at this meeting.
`
`TIEl Ad Hoc Massimo Sorbara, Globespan, reported that the ad hoc group is in the process of constructing a
`CAP/QAM specification. There is provisional agreements on the CAP and QAM sections. There is
`agreement on a system model. The draft document specifies bit sync, packet mode and A TM TCs, as well
`as an embedded operations channel. The group expects to have a complete document at the end of this
`week's meeting.
`
`TR30.1
`
`Les Brown presented the following liaison from TR30.1:
`
`T1E1.4/97-370 (Aspects ofxDSL Specifications which may affect V-Series Modems - TIA TR-30.1)
`
`Corrections: In the first line, "Recommendations" should have been "Standards".
`
`Discussion: It is important that xDSL devices not impact V-series modem performance. Note that modems
`currently use up to 3700 Hz. but implementations need some excess BW so the liaison says 4000 Hz.
`TR-30.l asks TIE1.4 to adopt the following requirements:
`
`Interference due to xDSL equipment should be limited to -75 dBm averaged over the 0-4000 Hz band
`•
`• POTS splitter LPF should introduce no significant attenuation or phase distortion out to at least 4 kHz
`
`The current Iss. 2 draft limits delay distortion to 200 s which is a typo, it should be 200 J.IS. V.34 can alter
`its symbol rate, but V.pcm uses a fixed symbol rate and needs the full 4 kHz. The impact from any given
`.. location will be the same all the time, and won't be random.
`
`Resolution: The Chair suggested that the WG make a "best effort" to limit impact on performance of
`V -series modems and take into account the concerns expressed in this liaison, with consideration for
`performance and implementability of both V-series and ADSL modems
`
`4. REVIEW/APPROVAL OF REVISED AGENDA
`
`The chair presented the following contribution:
`
`T1E1.4/97-268R2 (Revised Preliminary Agenda for this TIE1.4 Meeting - Chair)
`
`Discussion: The chair provided a revised agenda for the meeting.
`
`Resolution: Approved as modified by the WG. Due to the number of contributions, the WG agreed to a
`presentation time limit of 8 minutes. The WG agreed to a breakout session on ADSL Initialization and
`Timing on Monday evening, chaired by Peter Melsa.
`
`4
`
`

`
`TIE 1.4/97-362
`
`5. REVIEW/APPROVAL OF MINUTES OF JULY MEETING
`
`TlE1.4/97-261 The minutes of the interim meeting of June 30-July 2, which were distributed in the mailing,
`were approved with no changes.
`
`6. SPECTRALCOMPATmILITY
`
`TlE1.4/97-271 (On Spectral Compatibility with DFE-Based Systems using Baud Rate Sampling and Fixed
`Receive Filters - PairGain)
`
`Abstract: Spectral compatibility computations for DFE-based systems are normally based on an optimal
`DFE relation due to J. Salz, which implicitly assumes front-end filtering optimized for the SNR
`environment. Commonly deployed systems such as HSDL and ISDN, however, nominally employ fixed
`receive filters and baud rate sampling, which could result in significant deviations from optimal
`performance. This contribution modifies optimal DFE analysis to reflect the fixed receive filters of these
`systems. Margins are evaluated for HDSL, in the presence of echo-canceled ADSL and proposed HDSL2
`systems, showing substantial implementation dependence of results. Due to this dependence, an
`informative annex, ~ther than new normative text, is proposed for the Spectral Compatibility TR.
`
`Discussion: Some systems in the field have fixed received filters based on the noise environment expected
`at that time, but our spectral compatibility analysis has focused on DFE implementations. In baud sampled
`systems, noise is folded in non-coherently. Older HDSL systems have either weak analog filters (e.g. 3rd
`order at 200 kHz), while newer systems have higher order digital filters (e.g. can be modeled by 4th order at
`171 kHz. Older systems will exhibit degradation on the order of 2-3 dB due to the high frequency crosstalk
`from EC ADSL or HDSL2, but the impact on newer systems is minimal because of the greater high
`frequency attenuation of the integrated AFE. The contribution provides an analytical technique for looking
`at these systems.
`
`Resolution: There was support and no objections for including this method of calculation as an informative
`annex in the technical report. Contributions proposing draft text are needed.
`
`TlE1.4/97-336 (Interference from ADSL into HDSL due to fine PSD adjustments - ADTRAN)
`
`Abstract: Section 6.13.3 ofT1.413 and Section 6.15.3 of the Iss. 2 draft specifY the runtime PSD of the
`ATU-C transmitter. Although the average PSD over the entire used bandwidth is limited to -40 dBmIHz,
`the average PSD in smaller bandwidths is only limited by the peak PSD limitation of -36.5 dBmIHz • ....This--
`----eontribution shows that this may lead to reduced performance of existing services. An example is shown--- .-.
`where HDSL performance is reduced by 2.5 dB. Draft text, limiting the average PSD over an ADSL sub-
`band, is provided to eliminate this problem.
`
`Discussion: Frank Van der Putten said this scenario was possible but improbable, it is more likely that the
`higher tones would get the higher power. Peter Melsa thought the intent in T1.413 was to use gi only to
`equalize BER across the used tones for a given bi allocation. Two proposals were discussed:
`
`•
`•
`
`specifY total power below 300 kHz based on -38 and -40 dBmIHz psd
`specify 3 bands each 300 kHz wide
`
`Some members were concerned about extra rules for the gi algorithm to check (e.g. every time you do a
`bitswap.
`
`Resolution: For further study. Additional discussion was deferred to ADSL PSD on the agenda.
`
`5
`
`

`
`TIE 1.4/97-362
`
`TlE1.4/97-339 (The Importance of Testing Widely Deployed Equipment in the Presence ofInterference from New
`Line Codes - ADTRAN)
`
`Abstract: OPTIS includes a boost in a portion of the spectrum of the remote transmitter to achieve
`increased margins in the presence of ADSL NEXT. Study at ADTRAN has uncovered that the boost causes
`OPTIS to be spectrally incompatible with widely deployed HDSL hardware. Test results, using recent
`HDSL equipment from three North American manufacturers, show that OPTIS interference degrades the
`operating margins of HDSL by an average of 4 dB, with a minimum of 3 dB. These test results illustrate
`that simple theoretical calculations (based on optimum DPE receivers, utilizing ideal white noise) are not
`sufficient to ensure spectral compatibility with currently deployed equipment. The WG is encouraged to
`require such testing as a part of the HDSL2 line code selection process, particularly for those line codes
`employing spectral "boost", to ensure spectral compatibility with the installed base of loop technologies.
`
`Discussion: ADTRAN purchased HDSL equipment on the open market from 3 different vendors, and
`measured degradation due to crosstalk from OPTIS. Averaged across the vendors, there was 4 dB less
`margin with OPTIS crosstalk than with HDSL self-crosstalk. The authors recommend that the psd boost
`only be used if actual lab tests indicate it does not have an adverse effect.
`
`There was support for the concept of verifying that actual measurements agree with calculations, but some
`members preferred to resolve any difference by modifying the formulas to bring them in line with
`measurements. Several people liked the concept but thought the details would be difficult, e.g. specifying
`the crest factor of the noise. There was a suggestion that the proposed testing be used for WG selection of
`new line codes but not put in the spectral compatibility TR.
`
`Resolution: Placed on the spectral compatibility living list as Under Study. More implementation details
`are needed.
`
`TlE1.4/97-294 (Generic Approach and Common Specifications of Transmitter Power Spectral Density Masks for
`Twisted-Pair Loop Transmission Systems - Bellcore)
`
`Abstract: Bellcore has developed a generic approach and common specifications for transmitter PSD
`masks. These are to be used as guidelines to designers and service providers in the deployment of existing
`services or the development of new twisted-pair loop transmission systems. Specifically, the mask has taken
`into account existing systems such as ISDN DSL, HDSL, and DMT and CAP ADSL. To verify the
`appropriateness of the PSD masks, Bellcore has performed a bilateral analysis of the impact of the crosstalk
`of one system on another when they are collocated within the same binder group. This is compared to the
`impact ~ from the proposed PSD masks. Any xDSL system that complies with the PSD masks
`presented here could be deemed "spectrally compatible".
`
`Discussion: The objective is to propose psd masks for quickly gauging the compatibility of proposed new
`xDSLs. The compromise downstream mask is based on ISDN from 0-50 kHz, HDSL from 50-140 kHz,
`and ADSL from 140-1200 kHz (may have to be fine tuned depending on ADSL Iss. 2). Two upstream
`masks are presented, one for CSA loops, the other for RRDR loops.
`
`Gavin Young said he would like to see the impact ofEC DMT on HDSL and vice versa, also out-of-band
`masks considering impact on VDSL. George Zimmerman said a tweaked version of these masks would be
`better than the current placeholder for a simple test in 97 -180R 1. It was noted that real HDSL systems are
`designed around an 80 dB/oct rolloff, not 4th order filter on top of sinxlx, and that the assumption of 10th
`order filtering for HDSL repeaters does not conform to actual equipment.
`
`Resolution: Placed on the living list as Under Study. Contributions were requested for the next meeting to
`address comments brought up at this meeting.
`
`6
`
`

`
`TIE 1.4/97-362
`
`7. VDSL
`
`7.1 Service Characteristics & System Requirements
`
`TlE1.4/97-283 (Synchronizing Optical Network using GPS - Telia Research)
`
`Abstract: The Global Positioning System. GPS, can be used to synchronize ONUs in a VDSL system. It is
`possible to get a GPSIUTC time signal from commercially available GPS time reference receivers, with
`internal rubidium oscillator, that have an accuracy better than ±5 ns and a jitter that is less than 1 ns. This
`means that two ONUs synchronized with this type of GPS receivers will have less than 12 ns frame timing
`difference in the down-link data flow. If a GPS receiver with an ordinary high-stability OXCO is used the
`accuracy decreases to about 60 ns. This is still useable in this application.
`
`Discussion: The intent is to demonstrate that it is very feasible to synchronize multiple ONUs.
`
`Resolution: For information only.
`
`TlE1.4/97-284 (proposal: Additional bit-rate classes and granularity - Telia Research)
`
`Abstract: This contribution includes a proposal on additional bit-rates to be included in the VDSL
`standard. With some more data rates it is possible to introduce VDLS in an efficient way and enable
`flexibility for future not yet defined services and customer behavior. The lower data rates support operation
`over a little longer distances and gives a fall back solution for bad parts of the copper network.
`
`Discussion: It should be possible to start deploying VDSL from the CO, covering roughly half the
`customers with a 1500 m reach from the CO. Operators want the flexibility to select a lower rate and get
`extended reach. Note that Figure 1 is typical Swedish population distribution, and the distance is in meters.
`As ONUs are deployed, potential customers would be lost due to the coarse granularity in the bit rates
`available - for example, a customer needs 14 Mb/s but must move all the way up to 26 Mb/s. Initially, more
`and lower asymmetric and symmetric data rates are needed at the low end. In later stages of deployment,
`better granularity will be needed at the upper end. Lower data rates are needed as a fallback solution when
`severe crosstalk is encountered in the network.
`
`Hans Frizlen said the rate should be selectable for the purpose of adapting the NT to the LT, and extremely
`fine granularity is not needed - the author agreed. Tom Starr asked if we can pull in HDTV which is
`20 Mb/s, but it was noted that these are line rates, not payload rates.
`
`Resolution: For further study. (also see 97-342 below)
`
`TlE1.4/97-286 (Co-existence between VDSL and ADSL - Telia Research)
`
`Abstract: This contribution examines how ADSL and VDSL systems disturb each other when they share
`the same cable binder. Two VDSL scenarios are evaluated, CO based VDSL and cabinet based VDSL. The
`analysis shows that ADSL NEXT disturbs the VDSL upstream band significantly for a CO based VDSL
`scenario. For a cabinet based VDSL scenario, 20 VDSL distwbers can degrade ADSL capacity with 50%
`(VDSL NEXT) or with 20% (VDSL FEXT) depending on the chosen duplex scheme for VDSL.
`
`Discussion: CO based VDSL is heavily distwbed by ADSL NEXT (into VDSL upstream) and FEXT (into
`VDSL downstream). CO based ADSL is distwbed by NEXT and FEXT from cabinet based VDSL (50%
`reduction in ADSL throughput if NEXT and FEXT, or only 20% if VDSL designed so only FEXT disturbs).
`Gavin Young asked if the authors had looked at starting VDSL at 600 kHz instead of 200 kHz, to which
`they answered that they hadn't.
`
`Resolution: For information only.
`
`7
`
`

`
`TIE 1.4/97-362
`
`TlE1.4/97-324 (Effect of Bridged Taps at VDSL Frequencies - Lucent)
`
`Abstract: This contribution presents a study of the effects of bridged taps on VDSL transmission systems.
`Specifically, it is shown that the short bridged traps, which affect VDSL signals, are much more damaging
`than the longer bridged taps, which affect other xDSL signals, such as HDSL and ADSL. It is
`recommended that this issue be addressed in the VDSL system requirements document.
`
`Discussion: The first null occurs at fo(MHz) = 50/d(meters) = 164/f(feet) for polyethylene insulation. A
`long bridge tap is equivalent to introducing 3.5 dB of flat loss. Short bridge taps cause the first null to be
`deeper and wider, and therefore more damaging. The following conclusions are reached:
`
`•
`•
`•
`•
`
`bridge taps are much more damaging at VDSL frequencies than HDSL or ADSL
`they introduce significant linear distortion and overall propagation loss
`some combinations of 3 bridge taps can introduce overall propagation loss of up to 30 dB
`it is recommended that this issue be addressed in the VDSL system requirements document
`
`Tom Starr noted that this underscores the need for an active NT - Tom O'Shea pointed out the alternative
`of passive NT with UTP to the NID and coax in the home. Dick McDonald said that removing bridge taps
`in existing plant is difficult, but may be feasible for new plant.
`
`Resolution: The WG agreed to place in the system requirements document the information from this
`contribution allowing the degradation due to bridge taps to be calculated, plus an advisory note warning of
`the impact of bridge taps on reach.
`
`TlE1.4/97-319 (power Down in Multi-carrier Transmission - Alcatel)
`
`Abstract: When during data transport over the modem is interrupted for some time, the modem can go in
`a power down mode where its power consumption is only a fraction compared to normal operation. This
`contribution considers power down mode operation in multi-carrier systems where a fast transition from
`power down mode to normal operation mode is mandatory. In the described method, the transmission of all
`but one carrier is halted. The remaining modulated carrier ensures synchronization between transmitter
`and receiver. In addition, it provides a data channel enabling a fast power up.
`
`Discussion: Power consumption is a major issue for VDSL from the cabinet, on the order of 1.5 W per line.
`The traffic profile is characterized by periods of data traffic separated by long periods of inactivity. It is
`necessary to have a fast power up (a few ms), to maintain synchronization and framing, and to be able to
`measure SIN. The proposed method is to modulat.c QUI, ODe carrier (the pilot) to insure synchronization,
`adn to use the eoc to negotiate power down and request fast power up.
`
`John Cioffi pointed out that warm start is in the current system requirements document, but the warm start
`time is still tbd and under study.
`
`Resolution: The existing text in system requirements section 5.4 was changed from Under Study to
`Provisionally Agreed (no specific implementation agreed at this time) as an option. Making warm start a
`requirement was for further study. Warm start was also placed on the ADSL Iss. 3 living list as Under
`Study.
`
`TlE1.4/97-304 (Spectral Mask for VDSL with Power Boost - Analog Devices, Aware, BBT, Brooktree,
`Globespan, GTE Gov't Systems, Harris, Libit, Lucent, Metalink, Next Level, Orckit, Savan and Schmid)
`
`Abstnct: This contribution suggests a modified spectral mask in order to allow PSD boost in VDSL. In
`order to avoid increase of the maximal allowed total power relative to a non boosted PSD mask (~O
`dBmIHz), a combined PSD and total power limit is suggested.
`
`Discussion: The authors suggest a stairstep VDSL psd mask to concentrate power in the low frequency
`portion of the spectrum on long lines, taking advantage of higher cable balance. Once you build a system
`
`8
`
`

`
`TIE 1.4/97-362
`
`capable of transmitting a certain amount of total power, you might as well take advantage of that power on
`longer cables. The proposed maximum transmit power + 12 dBm, with notches 20 dB below the stairstep
`mask in the ham bands (note that the 18.068-18.168 MHz band was missed in the system requirements).
`The proposed masks would apply to both upstream and downstream.
`
`Jackie Chow agreed with the principal but said the proposal was line code dependent and made bit swap
`difficult - he preferred a single spec limiting total power. Several members supported a psd mask based on
`cable balance, plus a total power bound. J-F van Kerchkove suggested a smooth CUlVe rather than a
`stairstep. Les Humphrey noted that the proposed significant power boost from 1-3 MHz has potential for
`leaking into ADSL.
`
`Resolution: Placed on the living list as Under Study.
`
`TlE1.4/97-342 (Bit rate granularity for VDSL - Ericsson)
`
`Abstract: This contribution expresses Ericsson's view about the bit rate granularity needed in a general
`VDSL system. It is proposed that a defined minimum set of bit rates should be supported by the modem. It
`would then be up to each manufacturer to support more bit rates if they so wish. The extra granularity may
`not be mandatoI)', but a standard mechanism between the LT and NT must be defined so that
`interoperability is ensured if both modems support the same capability.
`
`Discussion: This is similar to Telia contribution 97-284. Ericsson is asking for a few more rates to achieve
`finer granularity, the exact rates are not critical, and Ericsson could agree to the rates in 97-284. Note that
`this contribution says "payload rate", 97-284 says "line rate", and the system requirements document says
`"payload rate".
`
`Resolution: The WG agreed to the principle of adopting rates similar to those in contributions 97-284 and
`97-342.
`
`T lEl.4/97 -131R2 (VDSL System Requirements)
`
`TlE1.4/97-010R2 (VDSL System Requirements Living List)
`
`Corrections: 97-010R2 has references to 97-133, which should be 97-131.
`
`Discussion: These 2 documents were not presented, but comments were invited from the WG. Peter Melsa
`raised the issue ofTI crosstalk, noting that Tl adjacent binder crosstalk-eould be a worse crosstalker than
`ADSL.
`
`Resolution: For information only. Peter Melsa agreed to bring a contribution at the next meeting
`expanding on his comment about Tl crosstalk.
`
`TlE1.4/97-165Rl (VDSL Editor suggestions for progress completion ofVDSL study phase)
`
`Abstract: The VDSL System Requirements is now in Revision 11. Some suggestions are made as to how
`to move this document to a completed state and to progress VDSL efforts in general. Specifically, 3
`champions are suggested in the areas ofT/C, POTSIISDN requirements, and Testing.
`
`Discussion: ETSI is moving fOlWard on VDSL, and has appointed champions. This contribution seeks
`volunteers to do the same in TlE1.4. The champions would coordinate activities on one of the 3 issues.
`
`Resolution: For information only. Volunteers should contact Tom Starr.
`
`9
`
`

`
`TIE 1.4/97-362
`
`TlE1.4/97-164RI (VDSL Editor's Suggestions to align VDSL system requirements with ETSI TM6 VDSL
`DTSITM)
`
`Abstract: The latest ETSI and ANSI VDSL requirements are largely consistent, but have a few deviations.
`Since TlE1.4 and TM6 have agreed to work closely on VDSL, the ANSI VDSL editor suggests the changes
`herein to the ANSI document to bring the two documents into sugstantial alignment. The editor suggests 3
`champions be appointed to pursue A TM and TIC specifications, POTSnSDN splitters, and Testing.
`
`Discussion: The editor drew attention to climatic requirements, psd, power down. operations and
`maintenance, and power back-off as areas where ETSI is significantly ahead of us. ETSI intends to finish
`in December, and John Cioffi suggested we appoint champions to talk to their champions. Hans Frizlen
`suggested bringing 97-164RI to the ETSI meeting the following week.
`
`Resolution: Neither Pete Chow nor Massimo SoIbara were able to attend the ETSI meeting, so Terry
`Throop volunteered and the WG agreed that he would represent TIE1.4 at that meeting, giving an oral
`report and bringing a copy of 97-164Rl as an informal liaison.
`
`7.2 Impairments & Loops
`
`TlE1.4/97-280 (RFl Characteristics of Category 5 Aerial Drop Wires - NORDXlCDT)
`
`Abstract: This contribution addresses the advantages of installing Category 5 aerial drop wires with
`respect to RFI.
`
`Discussion: The authors recommend cat 5 cable for rehab and green-field applications. It has 20 dB better
`FEXT and NEXT and significantly lower RFI. It allows DSL and Tl in the same cable, and provides 20 dB
`improvement in RFI when used for drop wire compared to F-drop.
`
`Resolution: For information only.
`
`TlE1.4/97-281 (RLGC and Crosstalk Parameters of Category 5 Outside Plant Cables - NORDXlCDT)
`
`Abstract: This contribution is a proposal to include RLGC parameters of Category 5 outside plant cable as
`part of the VDSL System Requirements Appendix.
`
`Discussion: NORDX would like the WG to include RLGC and crosstalk characteristics for this higher