TR-100 - Broadband Forum

TR-100 - Broadband Forum
TECHNICAL REPORT
TR-100
ADSL2/ADSL2plus Performance Test Plan
Issue: 1.0
Issue Date: March 2007
© The Broadband Forum. All rights reserved.
ADSL2/ADSL2plus Performance Test Plan
TR-100
Notice
The Broadband Forum is a non-profit corporation organized to create guidelines for
broadband network system development and deployment. This Technical Report has
been approved by members of the Forum. This document is not binding on the
Broadband Forum, any of its members, or any developer or service provider. This
document is subject to change, but only with approval of members of the Forum.
This document is provided "as is," with all faults. Any person holding a copyright in this
document, or any portion thereof, disclaims to the fullest extent permitted by law any
representation or warranty, express or implied, including, but not limited to,
(a) any warranty of merchantability, fitness for a particular purpose, non-infringement, or
title;
(b) any warranty that the contents of the document are suitable for any purpose, even if
that purpose is known to the copyright holder;
(c) any warranty that the implementation of the contents of the documentation will not
infringe any third party patents, copyrights, trademarks or other rights.
This publication may incorporate intellectual property. The Broadband Forum encourages
but does not require declaration of such intellectual property. For a list of declarations
made by Broadband Forum member companies, please see www.broadband-forum.org.
© The Broadband Forum. All rights reserved.
2
ADSL2/ADSL2plus Performance Test Plan
TR-100
Version History
Version
Number
1
Version Date
08 March 07
Version Editor
Herman Verbueken, Alcatel
Bell NV.
Changes
Issue 1
Editor:
Herman Verbueken
Alcatel Bell NV.
Copernicuslaan 50
B-2018 Antwerpen
Belgium
Tel. +32 3 240 7012
Fax +32 3 240 7798
herman.verbueken@alcatel-lucent.be
© The Broadband Forum. All rights reserved.
3
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table of Contents
1.
SUMMARY
7
1.1 Interoperability
7
1.2 Conventions
7
1.3 Scope
8
2.
REFERENCES
9
3.
TEST TOOLS REQUIREMENTS AND COMPENSATION
11
3.1 Accuracy of Loop simulators and noise sources
11
3.1.1 Loop Simulators
11
3.1.2 Noise Sources
17
3.1.3 Cabling
18
4.
COMMON TEST INFORMATION
20
4.1 Net data rates for DSLAMs
20
4.2 Compatibility Matrix/Definitions
20
4.3 Recording Temperature and Humidity
20
4.4 Sync State Definition
20
5.
EQUIPMENT FEATURES
21
5.1 DSLAM
21
5.2 CPE
21
6.
TEST CONFIGURATIONS
24
7.
PHYSICAL LAYER TEST CASES
27
7.1 Bitswap performance test
31
7.2 DSL Noise Spikes/Surges Tests
33
7.2.1 Isolated noise burst test
33
7.2.2 Repetitive High level Impulse Noise (REIN) test
34
7.2.3 Common mode noise susceptibility test
36
7.3 Stress Test
36
7.4 Verification of CRC error reporting by ATU-R (Basic CRC Functionality
Test) 37
7.5 Margin verification test procedure
38
8.
HIGHER LAYER TEST CASES
41
8.1 RFC2684 [11] Layer 3 Ethernet or USB Interface bridged mode
41
8.1.1 Packet Throughput Test
41
8.2 RFC2516 [12] PPPoE Throughput Test
44
8.3 RFC2364 [13] PPPoA End-to-End Connectivity Test
45
8.4 RFC2684 [11] IP Bridged End-to-End Connectivity Test
46
8.5 Power Cycle Test
47
ANNEX A PHYSICAL LAYER TEST CASES FOR SYSTEMS USING
G.992.3[1] AND G.992.5[2] ANNEX A.
48
48
A.1 North American Test set
© The Broadband Forum. All rights reserved.
4
ADSL2/ADSL2plus Performance Test Plan
TR-100
A.1.1 CPE Margin verification tests
A.1.2 Operation in the Presence of Impulse Noise Events
A.1.3 Verification of downstream bi/gi values
A.1.4 Loop tests with ports set for adaptive rate, automode operation
A.1.5 Loop tests with ports set for adaptive rate, single operating mode
A.1.6 Loop Tests with Ports Set For Fixed Rate
A.1.7 ADSL2plus Coexistence with Phoneline Networking
A.1.8 REIN noise impairment
71
A.2 European Test set
A.2.1 Noise and noise injection 72
A.2.2 CPE Margin verification tests
A.2.3 CO Margin Verification (Optional)
A.2.4 Verification of downstream bi/gi values
A.2.5 Loop Tests with Ports Set for Adaptive Rate
A.2.6 Loop Tests with Ports Set for Fixed Rate
A.2.7 Loop Tests with Ports Set for ADSL2 Annex L operating mode
A.2.8 REIN noise impairment
98
A.3 ADSL2plus European tests operating with DS-PSD mask above ADSL
A.3.1 Noise ADSL2plus over ADSL FD (rate adaptive)
A.3.2 Noise ADSL2plus over ADSL FD (fixed rate)
ANNEX B PHYSICAL LAYER TEST CASES FOR SYSTEMS USING
G.992.3/5 ANNEX B.
108
B.1 Annex B specific test setup information
B.1.1 Splitter model
B.1.2 Test loops / Loop characteristics
B.1.3 Noise and noise injection
B.1.4 Test procedure
B.2 Performance European Test set for ADSL2, Annex B
B.2.1 CPE Margin verification tests
B.2.2 CO Margin Verification (Optional)
B.2.3 Verification of Downstream bi/gi Values
B.2.4 Loop Tests with Ports Set for Adaptive Rate
B.2.5 Loop Tests with Ports Set For Fixed Rate
B.2.6 Performance in L2 mode
B.3 Performance Test set for ADSL2plus, Annex B
B.3.1 CPE Margin verification tests
B.3.2 CO Margin Verification (Optional)
B.3.3 Verification of downstream bi/gi values
B.3.4 Loop Tests with Ports Set for Adaptive Rate
B.3.5 Loop Tests with Ports Set For Fixed Rate
B.3.6 Performance in L2 mode
B.3.7 Performance with configured RFIBANDds
B.3.8 ADSL2plus European tests operating with DS-PSD mask above ADSL
ANNEX C PHYSICAL LAYER TEST CASES FOR SYSTEMS USING
G.992.3/5 ANNEX C
153
© The Broadband Forum. All rights reserved.
50
53
53
55
64
69
69
72
73
78
81
82
89
96
100
101
104
108
108
109
109
111
113
114
116
119
119
122
129
130
130
133
135
136
139
148
149
151
5
ADSL2/ADSL2plus Performance Test Plan
TR-100
ANNEX D ADSL2PLUS EQUIVALENT DISTURBER FREQUENCY
DOMAIN PROFILES OF GENERATORS G1 AND G2 FOR THE
EUROPEAN TEST SETS. 154
D.1 Equivalent disturber frequency domain profiles for Annex A.2 test set
154
D.2 Equivalent disturber frequency domain profiles for Annex A.3 test set
157
D.3 Equivalent disturber frequency domain profiles for Annex B.3 test set
158
APPENDIX I : EFFECT OF STATISTICAL VARIABILITY IN CPE
MANUFACTURING (INFORMATIVE)
160
APPENDIX II : COMPUTING THE MAXIMUM NUMBER OF FRAMES
PER SECOND (INFORMATIVE)
162
List of Figures
Figure 1/TR-100 : Amax vs. frequency for ADSL2plus annex-A.................................... 13
Figure 2/TR-100 : Amax vs. frequency for ADSL2plus annex-B ...................................... 13
Figure 3/TR-100: Test setup for loop tests for ADSL2/ADSL2plus external modems ... 24
Figure 4/TR-100: Test setup for throughput tests for ADSL2/ADSL2plus external
modems with Ethernet interfaces ............................................................................... 24
Figure 5/TR-100: Test set-up for USB modems.............................................................. 25
Figure 6/TR-100: Test set-up for internal modems .......................................................... 25
Figure 7/TR-100: Test set-up for European Annex B tests with splitters......................... 26
Figure 8/TR-100: Test setup for PPPoE testing with Ethernet modems .......................... 44
Figure 9/TR-100: Test setup for PPPoE testing with USB modems ................................ 44
Figure 10/TR-100: Test setup for PPPoA testing with Ethernet modems ........................ 46
Figure 11/TR-100: Test setup for PPPoA testing with USB modems.............................. 46
Figure 12/TR-100 Straight Loop Configuration ............................................................. 49
Figure 13/TR-100 Bridged Tap Loop Configuration ..................................................... 49
Figure 14/TR-100: G.pnt test setup .................................................................................. 71
© The Broadband Forum. All rights reserved.
6
ADSL2/ADSL2plus Performance Test Plan
TR-100
1. Summary
This document describes performance and interoperability test cases required for
ADSL2/ADSL2plus systems consisting of DSLAMs and CPE modems.
1.1
Interoperability
A CPE modem and a DSLAM are dynamically interoperable if they implement a
common and compatible set of features, functions and options and can demonstrate
satisfactory mutual communication in a real network architecture environment as
performance test conditions are varied and exercised. The term "compatible" is used to
mean that there are no conflicting requirements that will prevent the ADSL system from
achieving interoperability.
Systems are tested for Dynamic Interoperability on both standard loops and on a set of
additional loops. ADSL2/ADSL2plus Termination equipment (ATU-R and/or ATU-C)
will be required to be tested according to the tests stated in this document. An
interoperability statement with respect to this technical report is only applicable for ATUR/ATU-C combinations that have been tested against each other using the tests specified
in this document.
Throughout this document, the term “DSLAM” is understood to refer to the functionality
of the ATU-C. The terms “CPE”, “CPE modem” and “modem” are understood to refer
to the functionality of the ATU-R, unless stated otherwise. ATU-C functionality may be
provided by DSLAM units or digital loop carrier based (DLC) remote terminal units.
The wording “System under test” (SUT) is used when the behaviour is applicable to the
DSLAM and CPE combination.
1.2
Conventions
This document uses several words to signify the specification requirements. This section
defines these words as they shall be interpreted. The key words "must", "must not",
"required", "shall", "shall not", "should", "should not", "recommended", "may", and
"optional" in this document are to be interpreted as described below.
•
Must: This word, or the terms "required" or "shall", mean that the definition is
an absolute requirement of the specification.
•
Must Not: This phrase, or the phrase "shall not", means that the definition is an
absolute prohibition of the specification.
•
Should: This word, or the adjective "recommended", means that there may exist
valid reasons in particular circumstances to ignore a particular item, but the full
© The Broadband Forum. All rights reserved.
7
ADSL2/ADSL2plus Performance Test Plan
TR-100
implications must be understood and carefully weighed before choosing a
different course.
1.3
•
Should Not: This phrase, or the phrase "not recommended" means that there
may exist valid reasons in particular circumstances when the particular behavior
is acceptable or even useful, but the full implications should be understood and
the case carefully weighed before implementing any behavior described with
this label.
•
May: This word, or the adjective "optional", means that an item is truly
optional. One vendor may choose to include the item because a particular
marketplace requires it or because the vendor feels that it enhances the product
while another vendor may omit the same item. An implementation that does not
include a particular option must be prepared to interoperate with another
implementation which does include the option, though perhaps with reduced
functionality. In the same vein, an implementation that does include a particular
option must be prepared to interoperate with another implementation which
does not include the option (except, of course, for the feature the option
provides).
Scope
This test plan facilitates ADSL2/ADSL2plus over POTS and ISDN CPE / DSLAM
interoperability testing.
This test plan embodies operators' definitions of
ADSL2/ADSL2plus interoperability (between one DSLAM and one CPE at a time). The
test plan focuses on physical layer testing, and also validation and verification of selected
higher layer functionality. The test plan defines dynamic interoperability (performance)
as expected by leading carriers, specifying simulated network conditions under which
interoperability is required. The performance points in this test plan are based on ATU-C
equipment, capable of providing the maximum allowable power. ATU-C equipment
unable to provide this transmit power is considered to be out of the scope of this
interoperability testplan. The performance points may differ from the performance
requirements of G.992.3[1], G.992.5[2] and TS101388[8]. It does not replace operators'
pre-deployment testing.
This test plan defines tests for various physical layer functionalities and some higher
layer functionalities. A pass/fail indication result is provided for each functionality
tested.
Note: For ADSL1 operating mode, the reader is referred to TR-067[19]
Note: The Broadband Forum is developing a specification for functional
testing of ADSL2/2plus modems. The user of this document should be
advised that it is the intent to require that these functional tests be passed, as
well as the performance tests contained herein, in order to
claim
interoperability.
© The Broadband Forum. All rights reserved.
8
ADSL2/ADSL2plus Performance Test Plan
TR-100
2. References
Listed below are standards
interoperability test document.
referenced
throughout
this
ADSL2/ADSL2plus
[1] ITU-T Recommendation G.992.3 (01/2005), Asymmetric digital subscriber line
transceivers 2 (ADSL2), including amendment 1 (09/2005) and amendment 2
(03/2006).
[2] ITU-T Recommendation G.992.5 (01/2005), Asymmetric digital subscriber line
(ADSL) transceivers – extended bandwidth (ADSL2plus), including amendment
1 (07/2005) and amendment 2 (06/2006).
[3] ITU-T Recommendation G.996.1 (02/2001), Test procedures for digital subscriber
line (DSL) transceivers.
[4] ITU-T G.997.1 (06/2006), Physical Layer Management for Digital Subscriber Line
(DSL) Transceivers.
[5] ANSI T1.413 (1998), Network to Customer Installation Interfaces - Asymmetric
Digial Subscriber Line (ADSL) Metallic Interface.
[6] ANSI T1.417 issue 2 (09/2003), Spectrum Management for Loop Transmission
System
[7] ITU-T Recommendation G.117 (1996), Transmission aspects of unbalance about
earth.
[8] ETSI TS101388 V1.3.1 (2002-05), ADSL – European Specific Requirements.
[9] ETSI TS101952-1 V1.1.1 (04/2002) - Specification of ADSL splitters for European
deployment.
[10] IEEE 802.3u, Fast Ethernet.
[11] RFC2684, Multiprotocol Encapsulation over ATM Adaptation Layer 5 (AAL5).
[12] RFC 2516, Method for Transmitting PPP over Ethernet. (PPPoE).
[13] RFC 2364, PPP over AAL5 (PPPoA).
[14] ITU-T Recommendations G.9954 (02/2005) Phoneline networking transceivers –
Enhanced physical, media access, and link layer specifications
[15] Specification of the Access Network Frequency Plan applicable to transmission
systems connected to the BT Access Network", NICC Document ND1602:2005/08, Issue
3,
http://www.nicc.org.uk/niccpublic/Public/interconnectstandards/dsltg_spec/nd1602_2005_08.pdf
[16] ANFP issue 3 PSD masks spreadsheet
http://www.nicc.org.uk/niccpublic/Public/interconnectstandards/dsltg_spec/nd1602_2005_08.xls
[17] RFC 1242, Benchmarking terminology for network interconnection devices.
[18] RFC 2544, Benchmarking terminology for network interconnection devices (Test
methodology).
[19] TR-067 Issue2, Broadband Forum Technical Report “ADSL Interoperability Test
Plan”
© The Broadband Forum. All rights reserved.
9
ADSL2/ADSL2plus Performance Test Plan
© The Broadband Forum. All rights reserved.
TR-100
10
ADSL2/ADSL2plus Performance Test Plan
TR-100
3. Test Tools Requirements and Compensation
•
•
•
•
•
Loop simulator
Traffic simulator/analyzer with matching network interfaces
ATM switch/router
PC with USB/Ethernet interface
Noise sources for both ends of the line (loop simulator integral noise sources or
arbitrary waveform generators)
All these tools are part of configurations identified in figures 3, 4, 5 and 6. The ATM
switch/router and PC used for throughput testing must have adequate performance such
that they do not affect the measured throughput over the ADSL2/ADSL2plus link. The
ATM Switch or Simulator may be removed if traffic simulator/analyzer in use is capable
of terminating the ATM or IP traffic directly from the DSLAM.
3.1
Accuracy of Loop simulators and noise sources
3.1.1
Loop Simulators
a) Attenuation
North American region:
Loop attenuation which corresponds to the insertion loss is expressed in dB shall be
calculated from RLCG parameters using two-port ABCD modeling methodology as
specified in T1.417[6] Section B.3.1 (for both straight loops and loops with bridge taps).
The RLCG cable parameters shall be as specified in T1.417[6] Table B.2 "Cable model
parameters for 26-AWG twisted pair cable" and Table B.6 "Cable model parameters for
24-AWG twisted pair cable".
European region:
Loop attenuation, which corresponds to the insertion loss, is expressed in dB shall be
calculated from RLCG parameters using two-port ABCD modeling methodology as
specified in T1.417[6] Section B.3.1 (for both straight loops and loops with bridge taps).
The RLCG cable parameters shall be as specified in TS101388[8] Table A.1 "Parameter
set for generating distributed cable coefficients”.
For the loop simulator used in testing, the simulated loop attenuation shall be measured
over the frequency band [f1, f2], given by table 3-1 and 3-2 for the different annexes. At
least one measurement shall be made per fdelta interval. The Mean Error (ME) and Mean
Absolute Error (MAE) of the measured simulated loop attenuation values (in dB),
relative to the theoretical loop attenuation values (in dB), shall be calculated.
© The Broadband Forum. All rights reserved.
11
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 3-1/TR-100: Compensation boundaries for ADSL2
f1
f2
fdelta
Amax
Annex A/L
20kHz
1104kHz
10kHz
95dB
Annex B
120kHz
1104kHz
10kHz
95dB
Table 3-2/TR-100: Compensation boundaries for ADSL2plus
f1
f2
fdelta
Amax
Annex A
20kHz
2208kHz
10kHz
note1
Annex B
120kHz
2208kHz
10kHz
note2
Note 1: the Amax for ADSL2plus annex-A shall be the inverse function of the
ADSL2plus annex-A PSD mask, as specified in table 3-3
Note2: the Amax for ADSL2plus annex-B shall be the inverse function of the
ADSL2plus annex-B PSD mask, as specified in table 3-4
Table 3-3/TR-100: Definition of Amax for ADSL2plus Annex A
Frequency
25 kHz
1104 kHz
1622 kHz
2208 kHz
Level dB
95 dB
95 dB
85 dB
83.7 dB
© The Broadband Forum. All rights reserved.
12
ADSL2/ADSL2plus Performance Test Plan
TR-100
Figure 1/TR-100 : Amax vs. frequency for ADSL2plus annex-A
Table 3-4/TR-100: Definition of Amax for ADSL2plus Annex B
Frequency
120 kHz
1104 kHz
1622 kHz
2208 kHz
Level dB
95 dB
95 dB
85 dB
83.7 dB
Figure 2/TR-100 : Amax vs. frequency for ADSL2plus annex-B
© The Broadband Forum. All rights reserved.
13
ADSL2/ADSL2plus Performance Test Plan
TR-100
Mean Absolute Error (MAE) and Mean Error (ME) for loop X are given by Formula 1
and Formula 2:
Formula 1/TR-100
MAE loop X =
1
Ni
A −A
∑
{A A }
Ri
i∈
Ti
≤
Ti
+
Maxi
1
Nj
⎧
⎪
j∈⎨
⎪⎩
∑
ARj − AMax j
∑
ARj − AMax j
⎫
ATj > AMax j
⎪
⎬
0
.
5
−
<
−
ARj AMax j
⎪⎭
Formula 2/TR-100
ME loop X =
1
1
(
A
R
i − ATi ) +
∑
N i i∈{ATi ≤ AMax }
Nj
i
⎧
⎪
j∈⎨
⎪⎩
⎫
Max j
⎪
< −0.5 ⎬⎪
Max j
⎭
ATj > A
ARj − A
[positive error = too much attenuation]
ARi = Attenuation sample, in dB, of the measured loop X
ATi = Attenuation sample, in dB, of the theoretical loop X
The index “i” belongs to a set defined by the points necessary to measure the attenuation
in steps of fdelta or less and taking into account only those points between f1 and f2 for
which AT <= AMax dB.
Ni is the number of elements in the above set.
The index “j” belongs to a set defined by the points necessary to measure the attenuation
in steps of f or less and taking into account only those points between f1 and f2 for
delta
which A > A
T
Max
dB and A – A
R
Max
< -0.5 dB
N is the number of elements in the above set.
j
The loop simulator shall be compensated by adjusting the loop length such that the
absolute value of ME is minimized while maintaining an MAE less than 0.5 dB. This
accuracy requirement shall apply for all test loops.
b) Average noise floor
The average noise floor in the Wireline Simulator shall be lower than -150dBm/Hz
within the ADSL2/ADSL2plus band, measured on the device when powered on.
c) Impedance
North American region:
© The Broadband Forum. All rights reserved.
14
ADSL2/ADSL2plus Performance Test Plan
TR-100
Input impedance shall be calculated from RLCG parameters using two-port ABCD
modeling methodology as specified in T1.417[6] Section B.3.1 (for both straight loops
and loops with bridge taps).
The RLCG cable parameters shall be as specified in T1.417[6] Table B.2 "Cable model
parameters for 26-AWG twisted pair cable" and Table B6 "Cable model parameters for
24-AWG twisted pair cable".
European region:
Input impedance shall be calculated from RLCG parameters using two-port ABCD
modeling methodology as specified in T1.417[6] Section B.3.1 (for both straight loops
and loops with bridge taps).
The RLCG cable parameters shall be as specified in TS101388[8] Table A.1 "Parameter
set for generating distributed cable coefficients”.
Compensation method:
The impedance compensation shall be based on a difference in injected noise power
(captures the impact on the datarate).
The difference in injected noise power due to the variance of the input impedance of the
wireline simulator shall have a mean absolute error (MAE) of less than 0.5 dB from the
injected noise power using the theoretical input impedance, measured with the same
appropriate termination impedance in each case.
For North American noises the termination impedance is 100 ohms, for European noises,
this shall be according TS101388[8], section 5.1.
The difference in injected noise power is calculated in dB according to Formula 3.
Formula 3/TR-100
sim
loop
∆pi = 10 ⋅ log10 ( pout
) − 10 ⋅ log10 ( pout
)
= 10 ⋅ log10 (
Z inR, sim ( f i ) ⋅ Z L ( f i )
Z inR, sim ( f i ) + Z L ( f i )
2
) − 10 ⋅ log10 (
Z inR,loop ( f i ) ⋅ Z L ( f i )
Z inR,loop ( f i ) + Z L ( f i )
2
) dB
where
fi are the frequency bins.
The mean absolute error is defined in Formula 4.
Formula 4/TR-100
MAE (∆p ) =
1
N bins
∑ ∆p
i
i
© The Broadband Forum. All rights reserved.
15
ADSL2/ADSL2plus Performance Test Plan
TR-100
and the sum is over those bins in the passband where the insertion loss is less than 90 dB.
d) Phase
North American region:
Phase shall be calculated from RLCG parameters using two-port ABCD modeling
methodology as specified in T1.417[6] Section B.3.1 (for both straight loops and loops
with bridge taps).
The RLCG cable parameters shall be as specified in T1.417[6] Table B.2 "Cable model
parameters for 26-AWG twisted pair cable" and Table B6 "Cable model parameters for
24-AWG twisted pair cable".
European region:
Phase shall be calculated from RLCG parameters using two-port ABCD modeling
methodology as specified in T1.417[6] Section B.3.1 (for both straight loops and loops
with bridge taps). The RLCG cable parameters shall be as specified in TS101388[8]
Table A.1 "Parameter set for generating distributed cable coefficients”.
Mean Average Percentage Error for Phase delay shall be defined as in formula 5.
Formula 5/TR-100
where
Phase Delay(f) = unwrapped(phase(f))/ (2*pi*f)
f is the frequency
PDcable is the Phase delay for a theoretical loop, and
PDsim is the measured Phase delay for the simulator,
N is the number of frequencies used in the averaging.
Mean Average Percentage Error for Group Delay shall be defined as in formula 6.
Formula 6/TR-100
where
GDcable is the Group delay for a theoretical loop
GDsim is the measured Group delay for the simulator,
N is the number of frequencies used in the averaging.
© The Broadband Forum. All rights reserved.
16
ADSL2/ADSL2plus Performance Test Plan
TR-100
Points where |GDcable| is <= 0.1 microseconds shall not be included in the sum and N
shall be adjusted accordingly.
The maximum MAPE(PD) shall be 7%
The maximum MAPE(GD) shall be 7%
The measurement of the PD and GD used above shall made over a frequency range of ±
one bin, or 8.625kHz, starting at the lowest used channel frequency for the PDs in
question and ending at the channel below which EITHER the insertion loss exceeds 90dB
OR the frequency is the highest used frequency, whichever comes first.
The Group delay is defined using the formula 7.
Formula 7/TR-100
Where
phase is the unwrapped phase in radians
the difference in frequency between fi+1 and fi shall be 4.3125 kHz.
GD is not calculated or used at the two end frequency point.
3.1.2
Noise Sources
Each noise shall be measured independently at the ATU terminal. This shall be done for
one noise source at a time, using a zero-length loop. For North American cases both ATUs
are replaced by a 100 Ohm (±1%) resistor. For European cases the methodology in
TS101388[8] section 5.1.4.1 shall be used. The measured noise will be impacted by the
noise generator tolerance, the coupling circuit tolerance, cabling tolerance and noise
pickup.
For the noise source used in testing, the simulated noise level shall be measured over the
frequency band [f1, f2], given by table 3-5 for the different annexes. At least one
measurement shall be made per 10 kHz interval. The Mean Error (ME) and Mean Absolute
Error (MAE) of the measured simulated noise level values (in dBm/Hz), relative to the
theoretical noise level values (in dBm/Hz), shall be calculated.
Table 3-5/TR-100: Compensation frequency boundaries
f1
f2
Annex A
20 kHz
1104 kHz
Annex B
120 kHz
1104 kHz
© The Broadband Forum. All rights reserved.
17
ADSL2/ADSL2plus Performance Test Plan
TR-100
Mean Absolute Error (MAE) and Mean Error (ME) for noise X are given by formula 8 and
formula 9:
Formula 8/TR-100
MAE noise X =
1
M
∑P
Ri
⎧⎪
i∈⎨
⎪⎩
PTi
− PTi
⎫⎪
≥ −140 dBm / Hz ⎬
⎪⎭
Formula 9/TR-100
ME noise X =
1
M
∑ (P
Ri
⎧⎪
i∈⎨
⎪⎩
PTi
− PTi )
⎫⎪
≥−140 dBm / Hz ⎬
⎪⎭
[positive error = too much noise power]
PRi = power sample, in dBm/Hz, of the generated noise X
PTi = power sample, in dBm/Hz, of the theoretical noise X
The index “i” belongs to a set defined by the points necessary to measure the noise power
in steps of 10 kHz or less and taking into account only those points for which PTi >= 140 dBm/Hz.
M is the number of elements in the above set.
3.1.2.1
Noise Impairment Accuracy
All noise impairments used in this specification shall comply with the following
specifications. The theoretical noise level shall have a Gaussian amplitude distribution to 5
sigma.
The noise generator shall be compensated such that the absolute value of ME is minimized
while maintaining an MAE less than 0.5 dB.
Note: For noise calibration, there is measurement uncertainty that can not be compensated
for, consisting of the following contributions:
• absolute amplitude accuracy, vertical linearity and frequency response of the
measurement equipment used;
• tolerance of the calibration impedance.
3.1.3
Cabling
Cabling, switches and other equipment are needed to connect the DSLAM, the loop
simulator, the noise generator and the ATU-R. Care must be taken in order that the
minimum noise is coupled into this cabling, so the wiring should be kept short as
practically possible. Recommended cables should be Category 5 or better.
© The Broadband Forum. All rights reserved.
18
ADSL2/ADSL2plus Performance Test Plan
TR-100
For all loops with bridged taps at modem side, the modem interconnect cable shall be
included in the calculations. The modem cable used for these tests shall have an attenuation
of 6ft 26AWG.
For straight loops the length of this cable section is not important, as it is taken into
account during the compensation procedure.
Computer screens and power supplies radiate in the frequency bands used by
ADSL2/ADSL2plus. These devices should be placed at a distance from the setup or even
be switched off. This noise may be generated by either internal or external power supplies.
When the pickup noise levels are greater than -150 dBm/Hz, they will limit the
ADSL2/ADSL2plus performance and influence the test results.
The ATU-R and ATU-C and their wiring should be physically separated, since when
testing on long loops, crosstalk can occur between the cabling. Generally, starting from
attenuation levels of 70 dB and greater, care should be taken for the wiring to avoid
crosstalk.
To obtain the maximum accuracy the cables, switches and any other equipment used in the
link between the DSLAM and the Remote modem shall be contained within the
compensation process described above in section 3.1.1 (Loop Simulators).
© The Broadband Forum. All rights reserved.
19
ADSL2/ADSL2plus Performance Test Plan
TR-100
4. Common Test Information
4.1
Net data rates for DSLAMs
Throughout this document, the variable “net data rate” is used. This parameter is defined in
section 3.28 of G.992.3[1]. The “sync rate” is the net data rate achieved in showtime.
4.2
Compatibility Matrix/Definitions
A modem must achieve at least the minimum required performance in each test to claim
interoperability with the DSLAMs it is tested against. When a modem is tested against a
set of DSLAMs, it must achieve the minimum required performance in each test against
each DSLAM.
4.3
Recording Temperature and Humidity
The ranges of temperature and humidity of the test facility over the entire time of all the
tests herein shall be recorded in a manner similar to table 4-1. The acceptable range of
temperatures shall be between 15 °C/59 °F and 35 °C/95 °F. The humidity shall be
between 5% and 85%.
Table 4-1/TR-100: Temperature and Humidity
Parameter
High
Low
Temperature
Humidity
4.4
Sync State Definition
The modem sync state shall be defined as achieving showtime and capable of transferring
data.
© The Broadband Forum. All rights reserved.
20
ADSL2/ADSL2plus Performance Test Plan
TR-100
5. Equipment Features
The listed tables 5-1 and 5-2 shall be filled with the requested information before starting
the tests in order to have all the information about the EUT and to have a reproducible test
environment.
5.1
DSLAM
Table 5-1/TR-100: DSLAM Features (Informative)
Test Item
Results
DSLAM General Information
Vendor information (product name and revision)
HW Version
SW Version
Line Card Type, Version
Industry Standards Supported
Chipset (Vendor, HW and Firmware)
ADSL2/ADSL2plus Characteristics
supported max rates - downstream
supported max rates - upstream
possible coding options
Used duplex procedure (FDD, EC)
frequency usage (bin allocation) downstream
allowed usage Upstream bins (option below #33)
Support of extended framing parameters
Support of minimum INP control above INP=2
Dying Gasp detection
Power Cut Back implemented? (yes/no)
ATM Characteristics
Maximum Number of VCCs per DSLAM port
F5 OAM Support
VPI/VCI Ranges
Splitter Characteristics
Vendor information
Type (POTS, 2B1QISDN, 4B3T ISDN,…)
HW version
5.2
CPE
© The Broadband Forum. All rights reserved.
21
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 5-2/TR-100: CPE Features Tables (Informative)
Test Item
Results
CPE General Information
vendor information (product name and revision)
Industry Standards Supported
HW version
SW version
serial number
Modem form (interfaces)
PCI/USB driver version
Chipset (Vendor, HW and Firmware)
ADSL2/ADSL2plus Characteristics
supported max net data rate - downstream
supported max net data rate – upstream
possible coding options
used duplex procedure (FDD, EC)
allowed frequency usage downstream
used Upstream bins (option below #33)
Support of extended framing parameters
Support of minimum INP control above INP=2
Power Cut Back implemented? (yes/no)
dying gasp (yes/no)
ATM Characteristics
Maximum No. of VC’s
VPI/VCI Ranges
F4/F5 OAM Loopback, optional CC, AIS, RDI
supported QoS classes
ILMI supported (yes/no)
Protocols
RFC 2684 IP Bridging
RFC 2684 IP Routing
Bridge Filter
LLC-SNAP
VC-MUX
DHCP Client / Server
NAT
PAT
RFC 2364 PPPoA
RFC 2516 PPPoE
PAP / CHAP
Classical IP RFC 1577
Max number of active connections
© The Broadband Forum. All rights reserved.
22
ADSL2/ADSL2plus Performance Test Plan
Test Item
TR-100
Results
Other supported protocols
Splitter Characteristics
Vendor information
Type (POTS, 2B1QISDN, 4B3T ISDN,…)
HW version
© The Broadband Forum. All rights reserved.
23
ADSL2/ADSL2plus Performance Test Plan
TR-100
6. Test Configurations
Note: for Figures 3 through 7: high-impedance couplings may be integrated in noise
sources, and high impedance is defined as in G.996.1[3] Figure 3.
NOISE
SOURCE
NOISE
SOURCE
HI-Z
HI-Z
LOOP
SIMULATOR
DSLAM
CPE
MODEM
Figure 3/TR-100: Test setup for loop tests for ADSL2/ADSL2plus external modems
TRAFFIC
SIMULATOR/
ANALYZER
ATM SWITCH
OR
SIMULATOR
DSLAM
NOISE
SOURCE
NOISE
SOURCE
HI-Z
HI-Z
LOOP
SIMULATOR
CPE
MODEM
Figure 4/TR-100: Test setup for throughput tests for ADSL2/ADSL2plus external
modems with Ethernet interfaces
Note for Figure 4:
ƒ The ATM Switch or Simulator may be removed if traffic simulator/analyzer in use is
capable of terminating the ATM traffic directly from the DSLAM.
© The Broadband Forum. All rights reserved.
24
ADSL2/ADSL2plus Performance Test Plan
TR-100
PC
Noise Source
HI-Z
STM-1
STM-4
Ethernet
etc.
HI-Z
ATU-R
Ethernet
USB
Router
Line Simulator
ATU-C
Traffic routing
Traffic
generator/ analyzer
Figure 5/TR-100: Test set-up for USB modems
PC
Noise Source
HI-Z
HI-Z
ATU-C
Ethernet
Line Simulator
ATU-R
Router
STM-1
STM-4
Ethernet
etc.
Traffic routing
DSL line
Traffic
generator/ analyzer
Figure 6/TR-100: Test set-up for internal modems
Notes for Figure 5 and Figure 6:
ƒ
The PC must have a separate Ethernet interface or Ethernet card installed.
ƒ
The Ethernet card and the corresponding port on the traffic generator/analyzer
should be configured with IP addresses on the same network.
ƒ
The PC should be set-up to route traffic between the Ethernet interface and the USB
interface (see below). Note that the PPP session through the modem needs to be
initiated before this can be done, if PPP is to be used.
ƒ
With suitable choice of PC it is assumed that its effects on performance are
negligible.
© The Broadband Forum. All rights reserved.
25
ADSL2/ADSL2plus Performance Test Plan
TR-100
ƒ
The traffic generator/analyzer is used to measure end to end throughput, latency and
packet loss in exactly the same way as for other modem types (e.g., Ethernet).
ƒ
The PC setup examples below only apply to Windows PCs: (Note that the choice of
PC impacts performance and that this impact should be restricted.)
- On the PC, enable IP routing
-
Add a route on the PC to the traffic generator/analyzer port which is
connected to the router. Add a static route on the router to the Ethernet port of
the traffic generator/analyzer connected to the PC.
Noise
Lineport
Splitter
Lineport
Noise
Injection
ETSI Test Loop
Noise
Injection
Splitter
CPE
ADSL
-Port
TMNInterface
Noise
DSLAM
ISDNPort
ISDNPort
ISDN
Noise source
or ISDN
Impedance
ISDN
Noise source
or ISDN
Impedance
Figure 7/TR-100: Test set-up for European Annex B tests with splitters
© The Broadband Forum. All rights reserved.
26
ADSL2/ADSL2plus Performance Test Plan
TR-100
7. Physical Layer Test Cases
The loop simulators shall be calibrated relative to the nominal attenuation as defined in
chapter 3, section 3.1.1. Noise shall be injected through a high impedance network as
specified in G996.1[3], with simultaneous noise injection at both ends of the loop. The
noise injection shall be calibrated as defined in chapter 3, section 3.1.2.
Note: Although crosstalk models are intended for injection at a single end of the loop,
noise in this document is injected on both ends simultaneously, except for some testcases in
table A.1-13 and A.1-20, in order to reduce testing time. It is understood that noise levels
on short loops can be significantly higher.
Configuration is based on a set of common line settings as defined in table 7-1
General testprofiles define latency and INP settings as shown in table 7-2.
Specific testprofiles additionally define settings for DS/US net data rates, Operating
mode(ATSE) and RA-Mode. These setting are shown in table 7-3.
Deviations from these testprofiles and/or common line settings are indicated in the
description of each test or testsection.
Table 7-1/TR-100 Common Line Settings
Parameter
PMMode
Setting
all off
Latency/Bearer
Single
latency/one
Bearer
6 kbps
6 dB
standard value
no limitation or no additional power reduction in
at least 31dB
tests
0 dB
no influence on pass/fail criteria!
-40 dBm/Hz
G.992.3 default value
-38 dBm/Hz
G.992.3 default value
-40 dBm/Hz
G.992.3 default value
-38 dBm/Hz
G.992.3 default value
19.9 dBm
Annex-A G.992.3/5 default
(A.1.3.2)
19.3 dBm
Annex-B G.992.3/5 default
(B.1.3.2)
12.5 dBm
Annex-A G.992.3/5 default
(A.2.2.2)
13.3 dBm
Annex-B G.992.3/5 default
(B.2.2.2)
MSGMINds/us
TARSNRMds/us
MAXSNRMds/us
MINSNRMds/us
MAXNOMPSDds
MAXNOMPSDus
NOMPSDds
NOMPSDus
MAXNOMATPds
MAXNOMATPus
Description
no automatic transition to low power
states
Single latency path and Single Frame
Bearer operation
© The Broadband Forum. All rights reserved.
these
value
value
value
value
27
ADSL2/ADSL2plus Performance Test Plan
Parameter
CARMASKds
CARMASKus
PSDMASKds
RFIBANDSds
MAXBER
Setting
default
default
default
none
1E-7
TR-100
Description
Testcase dependant
Testcase dependant
only applicable for G.992.5
only applicable for G.992.5
Table 7-2/TR-100: General Testprofiles
General
Parameter
Setting
Testprofile
Low
delay Minimum INP 0 symbols
profile “F-1/0”
Maximum
S1
delay
Low
delay Minimum INP
profile “L-2/0”
Maximum
delay
High
delay Minimum INP
profile A “I16/2”
Maximum
delay
High
delay Minimum INP
profile B “I16/0.5”
Maximum
delay
0 symbols
2 ms
Description
No impulse noise protection
[G.997.1] 7.3.2.2 : The value
S1 indicates the Fast Latency
Path shall be used in the
G.992.1 operating mode and S
and D shall be selected such
that S ≤ 1 and D = 1 in ITU-T
Recommendations
G.992.2,
G.992.3, G.992.4, G.992.5 and
G.993.2 operating modes.
No impulse noise protection
One way interleaving delay
⎡S*D⎤/4
2 symbols
16 ms
One way interleaving delay
⎡S*D⎤/4
0.5 symbols
16 ms
One way interleaving delay
⎡S*D⎤/4
© The Broadband Forum. All rights reserved.
28
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 7-3/TR-100 : Specific Testprofiles
General
Specific Testprofile test
profile
ATSE
AU_RA_L_30000k
L-2/0
Automode
G.992.3 An.A
G.992.3 An.L AT_INIT
mask M1
G.992.5 An.A
30000-32
1048-32
AU_RA_I_30000k
I-16/2
Automode
G.992.3 An.A
G.992.3 An.L AT_INIT
Mask M1
G.992.5 An.A
30000-32
1048-32
A2_RA_F_16000k
A2_RA_I1/2_16000k
F-1/0
I-16/0.5
AT_INIT
AT_INIT
16000-32
16000-32
2016-32
2016-32
A2L_RA_I_16000k
I-16/2
A2_Fix_F_7288k
A2_Fix_I_7288k
A2_Fix_F_5952k
A2_Fix_I_5952k
A2_Fix_F_2400k
A2_Fix_I_2400k
A2_Fix_I_1200k
RAMode
DS net
US net
datarate(kbit/s datarate(kbit/s
) (max- min)
) (max-min)
AT_INIT
16000-32
2016-32
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
I-16/2
G.992.3 An.A
G.992.3 An.A
G.992.3 An.L,
mask M1
G.992.3 An.A
G.992.3 An.A
G.992.3 An.A
G.992.3 An.A
G.992.3 An.A
G.992.3 An.A
G.992.3 An.A
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
7288-7288
7288-7288
5952-5952
5952-5952
2400-2400
2400-2400
1200-1200
800-800
800-800
640-640
640-640
352-352
352-352
224-224
B2_RA_F_16000k
B2_RA_I_16000k
B2_Fix_F_7288k
B2_Fix_I_7288k
B2_Fix_F_3456k
B2_Fix_I_3456k
B2_Fix_F_864k
B2_Fix_I_864k
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
G.992.3 An.B
AT_INIT
AT_INIT
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
16000-32
16000-32
7288-7288
7288-7288
3456-3456
3456-3456
864-864
864-864
2016-32
2016-32
800-800
800-800
448-448
448-448
160-160
160-160
A2P_RA_F_30000k
A2P RA_I_30000k
A2P RA_I1/2_30000k
A2P_Fix_F_10000k
A2P_Fix_I_10000k
A2P Fix_F_7288k
A2P Fix_I_7288k
A2P Fix_F_5952k
A2P Fix_I_5952k
A2P_Fix_F_2400k
F-1/0
I-16/2
I-16/0.5
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
G.992.5 An.A
AT_INIT
AT_INIT
AT_INIT
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
30000-32
30000-32
30000-32
10000-10000
10000-10000
7288-7288
7288-7288
5952-5952
5952-5952
2400-2400
2016-32
2016-32
2016-32
832-832
832-832
800-800
800-800
640-640
640-640
352-352
© The Broadband Forum. All rights reserved.
29
ADSL2/ADSL2plus Performance Test Plan
TR-100
General
Specific Testprofile test
profile
ATSE
A2P Fix_I_2400k
A2P_Fix_F_600k
A2P Fix_I_600k
I-16/2
F-1/0
I-16/2
G.992.5 An.A MANUAL 2400-2400
G.992.5 An.A MANUAL 600-600
G.992.5 An.A MANUAL 600-600
352-352
128-128
128-128
B2P_RA_F_30000k
B2P RA_I_30000k
B2P_Fix_F_10000k
B2P_Fix_I_10000k
B2P Fix_F_7288k
B2P Fix_I_7288k
B2P Fix_F_3456k
B2P Fix_I_3456k
B2P_Fix_F_864k
B2P Fix_I_864k
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
F-1/0
I-16/2
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
G.992.5 An.B
2016-32
2016-32
832-832
832-832
800-800
800-800
448-448
448-448
160-160
160-160
RAMode
AT_INIT
AT_INIT
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
MANUAL
DS net
US net
datarate(kbit/s datarate(kbit/s
) (max- min)
) (max-min)
30000-32
30000-32
10000-10000
10000-10000
7288-7288
7288-7288
3456-3456
3456-3456
864-864
864-864
All operating modes are non-overlapped operation and only the mentioned operating mode
is enabled in to CO MIB.
The CO splitter used shall be the splitter integral to the DSLAM, if that option exists.
Otherwise, an external CO splitter as specified in Annex E of G.992.3[1] or ETSI
TS101952-1 [9] shall be used.
USB modems and NIC modems will be connected via an Ethernet card in a computer that
has the minimum configuration per the modem’s manufacturer.
An ATM or Ethernet switch may be included for DSLAM termination and IP return traffic.
General procedures for performance testing:
Tests will be performed at consecutive loops lengths identified in tables of the regionspecific annexes. The tests are initiated by placing the line from the DSLAM out of service
(OOS). Then the loop simulator is set with the appropriate noise impairments and loop
length, after which the line from the DSLAM is placed in-service (IS).
At each test point, the SUT must train within a total of
• 60 seconds, for single mode operation
• 120 seconds, for automode operation
Starting from the time that the ATU-C was placed IS. Following at least 60 seconds after
the train has occurred, the bi-directional net data rates and noise margins shall be recorded.
After it has reached its final train state, the modem shall remain in showtime for the
duration of the tests.
The line from the DSLAM shall then be placed OOS, the loop simulator loop length
incremented to the next test point loop length, the line from the DSLAM placed back IS,
© The Broadband Forum. All rights reserved.
30
ADSL2/ADSL2plus Performance Test Plan
TR-100
modem trained, and the data shall be recorded. This sequence shall continue until all loop
lengths defined in the table are complete.
The modem shall not be power cycled, rebooted or otherwise reinitialized between loop
length increments.
It is required that the SUT trains in every loop reach test. A failure to train in any test will
result in a failure of that section. This is required to eliminate the possibility of any modem
with interoperability ‘holes’ from passing the requirements contained within this document.
At times, the training process may result in a “False Train” condition. This situation occurs
when the modem initially indicates that it is trained but within a short time frame (<
allowed startup) goes into a retrain sequence.
If a false train occurs and the second train has not completed within the initial allowed
startuptime, a result of zero will be recorded into the result for that test point and the whole
section will be marked as a fail.
To obtain a result for each individual test, each test shall be performed once. In rateadaptive testing, any test point that fails to meet the net data rate requirement in
downstream direction by 96 kbps or less or in the upstream direction by 32 kbps or less
shall be re-tested 3 times. If a re-test is performed, then the maximum downstream value
achieved during testing, along with the associated upstream net data rate, shall be recorded.
If the SUT fails to sync within the allowed startup time, a result of zero will be recorded
into the result for that test point and the whole section will be marked as a fail.
ADSL2/ADSL2plus systems shall be tested according to tests specified for the different
regions and annexes. See table 7-4 for reference.
Table 7-4/TR-100: Pointer to physical layer testcases
ITU [1][2]
annex
TR-100 annex
7.1
Annex A/L
North
America
A1
Annex B
Annex C
Europe
Europe
Japan
A2/A3
B
C
Bitswap performance test
Table 7-5 describes a test that replicates the very rapid increases in amplitude of RFI that
occur on real circuits. When the RFI is applied it is expected the bits will be swapped away
from the tone affected by the RFI signal.
Note: Though the used signal has some similarity to RFI signals that can be
found in the field, the user of this test shall bear in mind that real narrow band
interference arising in the copper plant may significantly differ with respect to
its time and frequency statistical behavior.
© The Broadband Forum. All rights reserved.
31
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 7-5/TR-100: Bitswap performance test
Test
Configuration
[1] See Figure 3
[2] Configure the DSLAM for testprofile A2_RA_F_16000k or
B2_RA_F_16000k with the maximum downstream net data rate set
to 6432kbits/s and the maximum upstream net data rate set to
800kbits/s.
[3] Connect ATU-R and ATU-C with a stable connection at either
a. 9000 ft 26AWG (annex-A1)
b. 2.5 km ETSI loop #1 (annex-A2)
c. 2 km ETSI loop #1 (annex-B)
with –140dBm/Hz AWGN noise injected at both the ATU-C and
ATU-R ends.
[4] All single frequency tone amplitudes that are applied are referenced
in terms of power levels (dBm) at the injection point on the loop,
calibrated with the ATU-R and ATU-C modems replaced with
calibrated 100 Ohm ±1% resistors. Measurements performed into a
1kHz resolution bandwidth.
Note that with a 1kHz resolution bandwidth the power spectral
density value will be 30dB (in dBm/Hz) less than the power level (in
dBm), limited by the noise floor of the test equipment used for
calibration.
© The Broadband Forum. All rights reserved.
32
ADSL2/ADSL2plus Performance Test Plan
TR-100
of [1] Randomly select an integer value, n, in the range 70-100, ensuring
that the tone selected has assigned bits as described in the
downstream bits per tone map. Avoid the use of the pilot tone or any
unpopulated tones.
[2] Record and report the value of n used.
[3] Connect the ATU-R and ATU-C as per test configuration details.
The frequency of the interfering tone shall be set to n × 4.3125 kHz.
The power of the interfering tone shall be -110 dBm or less.
[4] Activate the management port to record the downstream bits per tone
map.
[5] Increase the power of the interfering tone to -50 dBm.
[6] Observe any downstream bit swap operation without retraining of the
modems or change in the downstream net data rate. After 2 minutes
verify the payload or CRC errors have stopped. (There may be some
initial errors seen when the signal is first inserted on the line.)
[7] Randomly select an integer value, n, in the range 10 – 20 for Annex
A or 40 – 50 for Annex B, ensuring that the tone n has assigned bits
as described in the upstream bits per tone map. Avoid the use of any
unpopulated tones.
[8] Record and report the value of n used.
[9] Connect the ATU-R and ATU-C as per test configuration details.
The frequency of the interfering tone shall be set to n × 4.3125 kHz.
The power of the interfering tone shall be -110 dBm or less.
[10] Activate the management port to record the upstream bits per
tone map.
[11] Increase the tone power to -50dBm.
[12]
Observe any upstream bit swap operation without retraining of
the modems or change in the upstream net data rate. After 2 minutes
verify the payload or CRC errors have stopped. (There may be some
initial errors seen when the signal is first inserted on the line.)
Expected Result The bit swap protocol re-deploys the allocation of bits among the subcarriers.
If the BER is not better than 1e-7 after 2 minutes, or a retrain occurred,
the Modem-DSLAM pair will be declared to fail the test.
Method
Procedure
7.2
DSL Noise Spikes/Surges Tests
The purpose of these tests is to verify that the xDSL functionality is not impacted by
sudden spikes or surges of noise on the line.
It also includes tests to verify that the xDSL functionality is not impacted by the presence
of repetitive impulse noise (REIN).
7.2.1
Isolated noise burst test
When an AWGN noise burst described in this section is applied to the modem under test
during Showtime, the modem shall not re-initialize within an observation period of 2
minutes. At most one test noise burst shall be applied within a 2 minute interval. The
noise shall be inserted at the customer-end of the line.
© The Broadband Forum. All rights reserved.
33
ADSL2/ADSL2plus Performance Test Plan
TR-100
Test settings are given in Table 7-6.
Table 7-6/TR-100: isolated noise burst test setting
Parameter
Value
Loop Length (26AWG)
3, 5, and 15kft
Target Noise Margin
6dB
Downstream/Upstream test profile AU_RA_L_30000k
At the end under test, bursts of band-limited white noise are coupled to the line using a
high-impedance coupler with nominal noise PSD levels as indicated below. These levels
are as measured into a 100Ω measuring set with another 100Ω in parallel.
The test noise burst shall be AWGN with amplitude of -86 dBm/Hz from 138kHz up to 2.2
MHz.
The out-of-band noise shall not be higher than –86 dBm/Hz.
A background noise level of -140dBm/Hz AWGN + 12 self-NEXT and FEXT simulated
noise shall be applied during modem start-up and throughout this test.
The burst length should be controllable with a resolution of ten milliseconds and the DSL
shall be monitored for re-initialization for 2 minutes after application of each burst. The
burst length changes by the loop length according to Table 7-7. The loops for these tests are
26 AWG.
Table 7-7/TR-100: isolated noise burst testloop and burst lenght
Loop Length (kft)
3
5
15
7.2.2
Burst length (msec)
4000
1000
100
Repetitive High level Impulse Noise (REIN) test
The REIN test shall use a “Burst of pseudo random AWGN” of 100µs duration at a level of
–90dBm/Hz differential mode. The repetition rate is defined in the regional annexes.
The system under test shall be tested for Interleaved rate adaptive and Interleaved fixed rate
profiles with an INP=2 and a delay=8ms.
The testprocedures are described in tables 7-8 and 7-9.
© The Broadband Forum. All rights reserved.
34
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 7-8/TR-100 : REIN test procedure Rate Adaptive mode
Test Configuration
Method of
Procedure
Expected Result
[1] The ATU-C and ATU-R are connected in turn through each loop as
specified in the REIN section of the regional annex A.1.8 for North
America and A.2.8 for Europe.
[2] The ADSL2/ADSL2plus profile shall be configured using the
Interleaved Rate Adaptive profile with INP=2 and delay=8ms.
as specified in the REIN section of the regional annex.
[3] The crosstalk noise impairment as defined in the REIN section of
the regional annex shall be applied at both ATU-C and ATU-R.
[4] The REIN noise impairment shall be applied at the ATU-R in
addition to the crosstalk noise.
[1] The target margin is set according to the table in the relevant
regional annex.
[2] The link is trained in the presence of the crosstalk noise and REIN
impairments.
[3] Wait for 3 minutes after initialization for bitswaps to settle.
[4] Record the net data rate R(kbps) and count the number of errored
seconds in next period of 2min.
[1] The broadband link shall operate in the presence of the REIN.
[2] If the link fails to train within 2min or the connection is dropped
before the end of the test, the result shall be declared a fail.
[3] The number of errored seconds measured in this 2 min period shall
be <= 1 for the test to pass.
[4] The expected results in the regional tables A.1.8 for North America
and in A.2.8 for Europe shall be met.
Table 7-9/TR-100 : REIN test procedure - Fixed Rate mode
Test Configuration
Method of
Procedure
[1] The ATU-C and ATU-R are connected in turn through each loop as
specified in the REIN section of the regional annex[2]. The
ADSL2/ADSL2plus profile shall be configured using the
Interleaved Fixed Rate profile with INP=2 and delay=8ms as
specified in the REIN section of the regional annex.
[2] The crosstalk noise impairment as defined in the REIN section of
the regional annex shall be applied at both ATU-C and ATU-R.
[3] The REIN noise impairment shall be applied at the ATU-R in
addition to the crosstalk noise.
[1] The link is trained in the presence of the crosstalk noise and REIN
impairments.
[2] Wait for 3 minutes after initialization for bitswaps to settle.
[3] Record the net data rate R(kbps) and count the number of errored
seconds in the next period of 2min.
© The Broadband Forum. All rights reserved.
35
ADSL2/ADSL2plus Performance Test Plan
Expected Result
7.2.3
TR-100
[1] The broadband link shall operate in the presence of the REIN.
[2] If the link fails to train within 2min or the connection is dropped
before the end of the test, the result shall be declared a fail.
[3] The number of errored seconds measured in this 2 min period shall
be <= 1 for the test to pass.
[4] The expected results in the regional tables shall be met.
Common mode noise susceptibility test
For further study.
7.3
Stress Test
The purpose of the test described in table 7-10 is to verify the stability of the CPE/Drivers.
1 individual test – 1 must be passed
Table 7-10/TR-100: Stress Test
[1] The test configuration shall be as shown in figures 4 to 6 in Section 6.
[2] The CPE may be set up for RFC 2684 bridging/routing, or PPP
bridging/routing.
[3] Configure the DSLAM with testprofiles A2_RA_F_16000k (Annex A) or
B2_RA_F_16000k (Annex B).
[4] Set up the loop simulators for either
a. 26 AWG at 12000 feet (annex-A.1).
b. ETSI loop#1 at 3600m (annex-A.2).
c. ETSI loop#1 at 2500m (annex-B).
On this loop, the modems are expected to train without excess margin.
[5] Inject –130 dBm/Hz white noise disturber at both ends of the loop.
Method
of [1] Train the CPE with the DSLAM.
[2] Wait for 1 minute after initialization.
Procedure
[3] Check reported margin and document as initial_reported_margin.
[4] Increase the noise level by MIN(initial_reported margin (dB) – 1 dB, target
margin (dB)) at CPE side.
[5] Configure the traffic generator/analyzer to provide MAC frames as a
payload source for the duration of the test. The engineer will need to adjust
the rate of the MAC frames to an acceptable level such that dropped frames
due to LAN-based collisions or otherwise are negligible. Record these rates
and the MAC frame size used for the test (suggested default 1024 bytes
including FCS).
[6] Run one over night BER test (8 hour minimum).
[1] The CPE shall not lose sync with the DSLAM at any time during the
Expected Result
overnight test. Any retrain will constitute a failure for this section.
[2] The BER shall be less than 1e-7 for the entire test.
Test
Configuration
Note: The Bit Error Rate (BER) will be approximated using the CRC error count
method as defined in section 7.5, Table 7-12 Fast path.
© The Broadband Forum. All rights reserved.
36
ADSL2/ADSL2plus Performance Test Plan
TR-100
7.4
Verification of CRC error reporting by ATU-R (Basic CRC Functionality
Test)
The purpose of this test described in table 7-11 is to verify that the CPE correctly reports
CRC errors. CRC error counts are the basis of margin verification tests; it is necessary to
verify if the CPE accurately counts and reports CRC errors.
Table 7-11/TR-100: Verification of CRC reporting by CPE
Test
Configuration
[1] The common line settings shall be according to Table 7.1.
[2] For Annex A.1 use the testprofile AU_RA_L_30000k.
For Annex A.2 use the test profile A2_RA_F_16000k and
A2P_RA_F_30000k.
For Annex B use the test profile B2_RA_F_16000k and
B2P_RA_F_30000k.
[3] Testloops for Annex A.1:
• 0.4 kft, -140 dBm/Hz AWGN
• 5 kft, NEXT – 5T1 (adj), -140 dBm/Hz AWGN
• 13 kft, NEXT – 24HDSL, -140 dBm/Hz AWGN
• 15 kft with 400 ft BT (24AWG), -140 dBm/Hz AWGN
Test loops for Annex A.2:
• 750m, noise FB ADSL2 impairment, A2_RA_F_16000k
• 4750m, -140 dBm/Hz AWGN, A2_RA_F_16000k
• 250m, noise FB ADSL2+ impairment, A2P_RA_F_30000k
• 1750m, -140 dBm/Hz AWGN, A2P_RA_F_30000k
Test loops for Annex B:
• 750m, noise FB ADSL2 impairment, B2_RA_F_16000k
• 3500m, -140 dBm/Hz AWGN, B2_RA_F_16000k
• 250m, noise FB ADSL2+ impairment, B2P_RA_F_30000k
• 1500m, -140 dBm/Hz AWGN, B2P_RA_F_30000k
Method
Procedure
Expected
Result
of [1] Connect CPE and DSLAM under the first required loop for the
regional annex and noise condition.
[2] Force a new initialization and wait for modems to sync. Wait for 2
minutes after initialization for bitswaps to settle.
[3] Force a “micro-interruption” of the loop at the CPE side with
duration of 10ms. Record the number of CRC errors reported. It
is expected that a micro-interruption will result in at least one
reported downstream CRC error.
[4] Repeat step 3 every 10 seconds, for a total test time of 120 seconds
(i.e. a total of 12 micro-interruptions are issued).
[5] Repeat steps 2 to 4 for all remaining testloops required for the
regional annex.
If each micro-interruption does not result in at least one reported
downstream CRC error, then the ATU-R has failed the basic CRC
error reporting test.
© The Broadband Forum. All rights reserved.
37
ADSL2/ADSL2plus Performance Test Plan
7.5
TR-100
Margin verification test procedure
The measurement time is based on the occurrence of 10 error events and a confidence
interval of 0.9.
With this confidence interval the required BER limit for a 1e-7 target bit error rate is 1.5e7.
To accelerate testing for some of the testcases, the desired number of observed error events
is reduced to a lower number, leading to higher required BER limits. When this is done, a
note is added under the table.
Also due to this confidence interval, one out of 10 margin verification test can result in a
false FAIL. Therefore when the first margin verification test fails, the test shall be redone
once, so that the confidence interval becomes 0.99.
Because of the significant dependency of achievable data rates on the noise margin, margin
verification tests are performed across several loop and noise scenarios of TR-100 to ensure
that there is no optimization of margin or modem performance for some specific test loops.
Table 7-12 shows how the BER shall be derived from the CRC count. Table 7-13 contains
the testprocedure on margin verification.
Table 7-12/TR-100: The equations for estimating BER
Modem
configuration
Fast
path
(maximum
delay = S1)
Interleaved
path
(maximum
delay ≠ S1)
Equations for estimating BER
BER =
Number of bit errors
15 * CRC_error_count
≅
Number of transmitted bits
data_rate * 1000 * test_time * 60
BER =
Number of bit errors
40 * CRC_error_count
≅
Number of transmitted bits
data_rate * 1000 * test_time * 60
Since CRC error counts are the basis of margin verification tests, it is necessary to verify if
the ATU-R accurately counts and reports CRC errors. A mandatory test procedure to verify
CRC error reporting is required and defined in section 7.4.
© The Broadband Forum. All rights reserved.
38
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 7-13/TR-100: Test procedure on Margin verification.
[1] The loop and noise conditions for the individual margin verification tests
are described in the regional annexes.
[2] The pass/fail criterion is the bit error ratio (BER) at the physical layer.
CRC error counts are used to estimate BER at the physical layer. Table 712 shows how the BER shall be derived from the CRC count.
of [1] Configure CPE and DSLAM according to the testprofile as indicated in the
regional annexes.
[2] Connect CPE and DSLAM to first test loop option, with the noise injected
at the specified reference power level. This power level is considered the 0
dB noise power level for that type of noise.
[3] Force a new initialization and wait for modems to sync.
[4] Wait for 3 minutes for bitswaps to settle.
[5] Check reported margin and document as initial_reported_margin.
[6] For CPE margin verification tests increase the noise power level by 1 dB at
CPE side only.
For CO margin verification tests increase the noise power level by 1 dB at
CO side only.
[7] Wait for 1 minute.
[8] Repeat steps 6 and 7 until the noise power is increased by
min(initial_reported_margin – 1, target margin) dB.
[9] At this point the power level of the noise is at the
min(initial_reported_margin – 1, target margin) dB level.
[10]
Execute a BER test for the duration as specified in each testcase.
Record the CRC and SES counts at the start and the end of the BER test.
Actual number of CRCs and SESs is the difference between these two
counts. Document the measured BER.
[11]
Repeat steps 2 to 10 for every test loop.
Test
Configuration
Method
Procedure
Expected
Result
In order for the BER result to be valid, the modems should not have reinitialized during steps 4 to 10.
If a re-initialization has occurred, the test for that loop option shall be repeated
once.
If the margin verification test case fails with a BER less than 2.5e-7 and higher
than the BER limit in the table, the test for that loop option shall be repeated
once.
The CPE modem passes the margin verification test, if
[1] for every test loop: Estimated BER is less than the limit value in the
margin verification table of that loop option
[2] No SES has been reported
An
explanation
of
the
regional
margin
verification
tables
is
given
below.
Each test is specified by the type and level of the injected noise and by the type and length
of the loop.
The title of each test specifies the type and level of the injected noise, the loop option and
the used testprofile.
Column 1 of each table specifies the loop type (e.g. 26AWG) and loop length (e.g. 3kft).
© The Broadband Forum. All rights reserved.
39
ADSL2/ADSL2plus Performance Test Plan
TR-100
Column 2 specifies the required test time in order to observe approximately 10 CRC error
events at a target BER of 1e-7. These test times are calculated based on the data rates in the
corresponding performance requirements section. To allow for modem connections with
slightly lower than the anticipated data rates, the test durations are rounded up in
increments of 5 minutes.
Column 3 It is required that the achieved DS(US) net data rate shall be recorded twice:
first at the start of the margin verification test and also at the end of the margin verification
test. This captures a potential modem retrain and connection at lower DS net data rates due
to the increase in injected noise level.
It is expected that the actual datarate is close to the anticipated one. If the actual datarate is
significantly lower then the anticipated one, the test results may not be valid.
The actual net data rate is not a pass/fail criteria of this test.
Column 4 asks for the insertion of the measured DS(US) CRC count after the injected
noise level has been increased by min(initial_reported_margin – 1, target margin) dB.
Measurement of initial_reported_margin shall be done from the DSLAM.
Column 5 asks for the computation of the estimated BER from the number of observed
CRC error events according to Table 7-12. If the estimated BER is smaller than the BER
limit in this column, then the test PASSES, else it FAILS.
Column 6 asks for the insertion of “PASS” or “FAIL”.
© The Broadband Forum. All rights reserved.
40
ADSL2/ADSL2plus Performance Test Plan
TR-100
8. Higher Layer Test Cases
8.1
RFC2684 [11] Layer 3 Ethernet or USB Interface bridged mode
The throughput tests specified satisfy the terminology criteria of RFC 1242 [17] and the
test methodology specified in RFC 2544 [18].
• From RFC 1242 [17], throughput is “the maximum rate at which none of the offered
frames are dropped by the device.”
• From RFC 2544 [18], the methodology to measure throughput is to “Send a specific
number of frames at a specific rate through the DUT and then count the frames that
are transmitted by the DUT.”
• If the count of offered frames is equal to the count of received frames, the rate of
the offered stream is raised and the test is rerun. If fewer frames are received than
were transmitted, the rate of the offered stream is reduced and the test is rerun.
In this test specification, the number of frames transmitted from one end of the
ADSL2/ADSL2plus link is compared with the number of frames received at the far end of
the ADSL2/ADSL2plus link. The method used for computing the maximum number of
frames per second is given in Appendix II.
8.1.1
Packet Throughput Test
The purpose of the test contained in table 8-2 is to verify the throughput for a selected list
of provisioned line rates (down/up) using IP Frame transfers of varying length.
The packet or frame size is defined in RFC 1242 [17], section 3.5, Data Link Frame Size,
and states “The number of octets in the frame from the first octet following the preamble to
the end of the FCS, if present, or to the last octet of the data if there is no FCS”. This
specifically defines the packet or frame size in terms of the MAC frame, and hence is not
the IP packet size.
Throughput shall be measured for MAC frame length (see RFC 1242 [17] Section 3.5).
The packet throughput requirements in the tables 8-3, 8-4 and 8-5 includes trailer overhead
for AAL5 CPCS-PDU without preservation of the Ethernet FCS.
The DSLAM is configured similar to the fixed profile A2P_Fix_F_10000k, but with the
bitrates configured as in table 8-1.
Table 8-1/TR-100: Packet Throughput Test bitrates
DS bitrate (kbps)
US bitrate (kbps)
12000
800
18000
800
24000
800
© The Broadband Forum. All rights reserved.
41
ADSL2/ADSL2plus Performance Test Plan
TR-100
42 individual tests – 42 tests must be passed
Table 8-2/TR-100: Packet Throughput Test
[1] The configuration shall be as shown in figures 4 to 6,
depending on the customer interface of the modem.
[2] Set up the loop simulators for a loop of 2kft 26 AWG cable,
or a loop of 600m 0.4 mm cable, with white noise (–
140 dBm/Hz) injected at both the upstream and downstream
ends of the loop.
[3] Setup the traffic generator/analyzer to send frames in one
direction with destination MAC address equal to the unicast
source MAC address of frames sent in the opposite direction.
[4] Set the CPE modem for bridged mode with LLC
encapsulation.
Method of Procedure [1] Configure the DSLAM similar to the fixed profile
A2P_Fix_F_10000k, but with the bitrates configured as in
table 8-1 for each of the throughput tables 8-3 to 8-5.
[2] Let the CPE train.
[3] Setup traffic generator/analyzer to perform throughput test
for selected frame length and connect rate.
[4] Set the throughput rate of the upstream direction to 50% of
the maximum theoretical value sustainable by the
ADSL2/ADSL2plus net data rate.
[5] Test for the throughput in the downstream direction. Record
the downstream throughput rate as frames per second. The
test shall be run for 60 seconds.
[6] Set the throughput rate of the downstream direction to 50%
of the maximum theoretical value sustainable by the
ADSL2/ADSL2plus net data rate.
[7] Repeat the test for throughput in the upstream direction.
Record the upstream throughput rate as frames per second.
The test shall be run for 60 seconds.
[8] Divide the analyzer frames per second by the maximum FPS
for the connect rate and frame size.
[9] Record as percentage of maximum connect rate.
Expected Result
Based on the throughput tables: The percentage of frames
achievable for all DSL modems (Ethernet and USB) shall be
85%.
Test Configuration
Note: This 85% also assumes the downstream connect rate does not exceed the
USB bus maximum. In that case, the USB bus maximum will dictate the Max
FPS recorded
© The Broadband Forum. All rights reserved.
42
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 8-3/TR-100:Throughput Test Results: Connect Rates DS: 12000 kpbs US: 800
kbps.
Analyzer Recorded FPS
Frame
Size
64
128
256
512
1024
1280
1514
DS
US
Max FPS
DS
US
14150
7075
4716
2358
1286
1010
884
943
471
314
157
85
67
58
% of Max
DS
US
Pass/Fail
(85% limit)
DS
US
Table 8-4/TR-100:Throughput Test Results: Connect Rates DS: 18000 kbps US: 800
kbps.
Analyzer Recorded FPS
Frame
Size
64
128
256
512
1024
1280
1514
DS
US
Max FPS
DS
US
21226
10613
7075
3537
1929
1516
1326
943
471
314
157
85
67
58
% of Max
DS
US
Pass/Fail
(85% limit)
DS
US
Table 8-5/TR-100: Throughput Test Results: Connect Rates DS: 24000 kbps US:
800kbps.
Analyzer Recorded FPS
Frame
Size
64
128
256
512
1024
1280
1514
DS
US
Max FPS
DS
US
28301
14150
9433
4716
2572
2021
1768
943
471
314
157
85
67
58
% of Max
DS
© The Broadband Forum. All rights reserved.
US
Pass/Fail
(85% limit)
DS
US
43
ADSL2/ADSL2plus Performance Test Plan
8.2
TR-100
RFC2516 [12] PPPoE Throughput Test
Table 8-6 describes the testprocedure for the PPPoE packet throughput test
PPPoE Session
PC
Noise Source
HI-Z
HI-Z
ATU-R
Ethernet
STM-1
STM-4
Ethernet
etc.
Line Simulator
Router
(PPPoE
Server)
ATU-C
Figure 8/TR-100: Test setup for PPPoE testing with Ethernet modems
PPPoE Session
PC
Noise Source
HI-Z
HI-Z
ATU-R
USB
STM-1
STM-4
Ethernet
etc.
Line Simulator
Router
(PPPoE
Server)
ATU-C
Figure 9/TR-100: Test setup for PPPoE testing with USB modems
2 individual tests – 2 tests must be passed
© The Broadband Forum. All rights reserved.
44
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table 8-6/TR-100: PPPoE Packet Throughput Test
Methodology
Test Configuration
Method of
Procedure
Expected Result
8.3
The throughput, at the PPPoE layer, is assessed by measurement
at the throughput at the TCP layer. The ratio, of the throughput
at the TCP layer to the data rate made available to ATM cells, is
determined. This ratio is then compared to a figure that is
slightly below the ideal value of 85.4%.
The frame size indicated below is the length of the IP frame.
[1] The configuration shall be as shown in figure 8 or 9,
depending on the customer interface of the modem.
[2] Set up the loop simulators for a NULL loop.
[3] The buffer size (used by the measurement software) must be
at least 64 kB
[4] Configure the DSLAM with testprofile AU_RA_I_30000k
and the maximum delay set to 8 ms
[5] Set the CPE modem for PPPoE mode.
[1] Using software capable of measuring throughput at the TCP
layer, initiate a download, in the downstream direction, of a
file having length of 50 Mbytes. Complete 10 such
transactions.
[2] Record the average throughput rate result in units of bits per
second.
[3] Calculate the ratio of the downstream throughput to the
downstream date rate.
[4] Using software capable of measuring throughput at the TCP
layer, initiate an upload, in the upstream direction, of a file
having length of 1.0 Mbytes.
[5] Record the average throughput rate results in units of bits per
second.
[6] Calculate the ratio of the upstream throughput to the
upstream date rate.
[1] In both the upstream and downstream directions, the ratio
shall equal or exceed 0.84.
RFC2364 [13] PPPoA End-to-End Connectivity Test
Table 8-7 describes the testprocedure for the PPPoA end-to-end connectivity test.
1 individual test – 1 must be passed
© The Broadband Forum. All rights reserved.
45
ADSL2/ADSL2plus Performance Test Plan
TR-100
PPPoA Session
PC
Noise Source
HI-Z
HI-Z
ATU-R
Ethernet
Line Simulator
STM-1
STM-4
Ethernet
etc.
Broadband
Access
Server
ATU-C
Figure 10/TR-100: Test setup for PPPoA testing with Ethernet modems
PPPoA Session
PC
Noise Source
HI-Z
HI-Z
ATU-R
USB
STM-1
STM-4
Ethernet
etc.
Line Simulator
Broadband
Access
Server
ATU-C
Figure 11/TR-100: Test setup for PPPoA testing with USB modems
Table 8-7/TR-100: PPPoA
Test Configuration [1] See Figure 10 or 11 (as applicable)
Method
of [1] Terminate PPPoA session between the computer and
the broadband access server.
Procedure
[2] Verify connectivity by passing traffic over this PPPoA
session.
[3] Tear Down the PPPoA session.
[1] Transmitted packets are received.
Expected Result
[2] The PPPoA session has been torn down correctly.
8.4
RFC2684 [11] IP Bridged End-to-End Connectivity Test
Table 8-8 describes the testprocedure for the IP Bridged connectivity test.
© The Broadband Forum. All rights reserved.
46
ADSL2/ADSL2plus Performance Test Plan
TR-100
1 individual test – 1 must be passed
Table 8-8/TR-100: Verify IP Bridged RFC 2684 [11]
[1] See Figure 4, applicable to Ethernet modem only
Test Configuration
Method
of [1] Configure the test environment including the ATU-R and
the computer so that the ATU-R/CPE Ethernet port
Procedure
terminates a bridge section.
[2] The second termination of the bridge section should be
implemented at an appropriate device within the testing
environment (e.g. DSLAM, PoP).
[3] Pass IP packets over the complete bridge section and
verify the proper reception at the destination (e.g. PoP,
Host PC).
Expected Result
8.5
[1] Transmitted packets are received.
Power Cycle Test
Table 8-9 describes the testprocedure for the power cycle test.
1 individual test - 1 must be passed
Table 8-9/TR-100: Power Cycle Test
The configuration shall be as shown in Figures 5 - 7,
depending on customer interface of modem (no loop
length, bridge tap nor noise injected)
Method of Procedure [1] Connect Modem to test setup.
[2] Configure
the
DSLAM
with
testprofile
AU_RA_I_30000k and the maximum delay set to 8
ms
[3] Train modem.
[4] Follow procedure to establish a data link, if
necessary.
[5] Send arbitrary data (verify received).
[6] Disconnect power from modem for 30 seconds.
[7] Allow modem to train.
[8] Follow procedure to re-establish the data link.
[9] Re-connect power cable.
[10]
Send arbitrary data (verify received).
Test Configuration
Expected Result
Modem re-powers.
passes data.
Link re-covers and modem
© The Broadband Forum. All rights reserved.
47
ADSL2/ADSL2plus Performance Test Plan
TR-100
Annex A Physical Layer Test Cases for systems using
G.992.3[1] and G.992.5[2] Annex A.
A.1
North American Test set
Listed are three known sources of variability that need to be taken into account when
verifying the interoperability of a CPE and DSLAM combination.
1. CPE Variability
For a modem type to be considered compliant, unit(s) submitted for compliance
testing shall pass the performance requirements in this section. Taking into account
the statistical variability in the manufacturing process, it is expected that the large
majority of randomly-selected units will pass these requirements. (See Appendix I)
2. Test Environment Variability
The rate requirements in this section take into account a 0.5dB variability in the test
environment.
3. Bridged Tap Noise Injection Variability
The noise injection method of the present version of the document for bridged tap
sections may lead to:
• results differing from those on real loops,
• lower repeatability of results (due to, for example, varying noise floor and
impedance) from test environment to test environment even if they are
compliant to this document.
The impact on the rate requirements contained herein and ways to improve test
conditions are under study. A future version of this document may address this
issue.
Tests shall be performed according to the general procedure described in section 7.
CO and CPE ends shall operate over POTS, using the FDM spectral mask.
The ADSL2plus near-end and far-end self-crosstalk (NEXT and FEXT) shall be based on
the template PSD masks from G.992.5[2] and the loop transfer functions from T1.417[6]
• NEXT shall be calculated using the “Simplified Next Equation” from T1.417[6]
section A.3.2.1.1. and G.992.5[2] PSD for the specified loop configuration, gauge,
length and number of disturbers.
• FEXT will be calculated using the “Far end crosstalk” equation from T1.417[6]
section A.3.2.2 and the G.992.5[2] PSD for the specified loop configuration, gauge,
length and number of disturbers.
Following loops shall be used (T1.417[6] cable models) :
• 26awg variable length straight loop
© The Broadband Forum. All rights reserved.
48
ADSL2/ADSL2plus Performance Test Plan
•
TR-100
26awg loop with 24awg bridge tap segment at the customer end of the loop (at CPE
side)
Straight loop tests shall use the loop configuration shown in Figure 12. A 0ft bridge tap
may be attached to the ATU-R end of the straight loop during testing. Bridged Tap loops
shall use the loop configuration shown in Figure 13.
ATU-C
ATU-R
26 AWG
0-20000 ft
Figure 12/TR-100 Straight Loop Configuration
24 AWG
0 - 1500 ft
ATU-C
ATU-R
26 AWG
0-17500 ft
Figure 13/TR-100 Bridged Tap Loop Configuration
All tests for section A.1 shall be executed with configured RFI band notching.
Table A.1-1 shows the HAM band that shall be notched. Other HAM bands may or may
not be notched as well.
The Ham band is specified according G.997.1 section 7.3.1.2.10
‘RFI bands
(RFIBANDSds)’, using two breakpoints [ti;ti+1], corresponding to the low level of the
notch.
These breakpoints are referred to as points t(i+2) and t(i+3) in G.992.5 section 8.5.1 item 5
‘RFI band specification’ (where the index i represents the i-th RFI band)
Table A.1-1/TR-100 Ham band specification
Band start
kHz
1800
Band stop
kHz
2000
© The Broadband Forum. All rights reserved.
PSD level
(dBm/Hz)
-80
49
ADSL2/ADSL2plus Performance Test Plan
TR-100
Note: Modems passing this test are not necessarily meeting the spectrum
compatibility as defined by T1.417 for North America, or any other similar
requirement specification. Users of this document should carry on additional
testing to ensure spectral compatibility before deployment.
A.1.1 CPE Margin verification tests
Section 7.5 gives a description and testprocedure for the margin verification tests in tables
A.1-2 up to A.1-8.
3
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Estimated BER,
based on equations
in Table 7-12,
interleaved path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
Pass / Fail
Anticipated DS
net data rate:
>= 12254 kbps
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1-2/TR-100: Margin Verification (3000 ft 26awg, 12 ADSL2plus DN Fext, 140dBm/Hz AWGN), testprofile AU_RA_L_30000k
10
5
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Achieved DS
net data rate
(test end)
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated BER <
1.5e-7
then PASS
else FAIL
Pass / Fail
Anticipated DS
net data rate:
>= 9648 kbps
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1.3/TR-100: Margin Verification (5000 ft, NEXT – 5 T1 adjacent, 140dBm/Hz AWGN) , testprofile AU_RA_L_30000k
10
© The Broadband Forum. All rights reserved.
50
ADSL2/ADSL2plus Performance Test Plan
TR-100
12
with 750ft
BT
(24AWG)
Anticipated DS
net data rate:
>= 3842 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
Pass / Fail
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1-4/TR-100: Margin Verification (12000 ft, 750 ft BT (24AWG), -140dBm/Hz
AWGN), testprofile AU_RA_L_30000k
55
13
Anticipated DS
net data rate:
>= 1624 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Achieved DS
net data rate
(test end)
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Pass / Fail
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1-5/TR-100: Margin Verification (13000 ft, NEXT - 24 HDSL, -140dBm/Hz
AWGN), testprofile AU_RA_L_30000k
45
© The Broadband Forum. All rights reserved.
51
ADSL2/ADSL2plus Performance Test Plan
TR-100
15 with
400ft BT
(24WG)
Anticipated DS
net data rate:
>= 1252 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
Pass / Fail
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1-6/TR-100: Margin Verification (15000 ft, 400 ft BT (24AWG), -140dBm/Hz
AWGN), testprofile AU_RA_L_30000k
55
17
Anticipated DS
net data rate:
>= 799 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Achieved DS
net data rate
(test end)
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated_BER <
1.75e-7
then PASS
else FAIL
Pass / Fail
Test time (minutes)
Loop length
(kft) 26AWG
Table A.1-7/TR-100: Margin Verification (17000 ft, -140dBm/Hz AWGN), testprofile
AU_RA_L_30000k
45
Note: Due to the low, achievable DS data rates under this loop and noise
condition, the observation of 10 error events would require approximately 85
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of observed
error events is reduced to 5. To remain consistent with previous confidence
levels of estimated BER, the range of allowed estimated BER is increased from
1.5-e7 to 1.75e-7
© The Broadband Forum. All rights reserved.
52
ADSL2/ADSL2plus Performance Test Plan
TR-100
0.4
A.1.2
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
min(initial_reported_margin
– 1, target margin) dB
Achieved DS
net data rate
(test end)
Estimated BER,
based on equations in
Table 7-12,
interleaved path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Pass / Fail
Anticipated DS
net data rate:
>= 23768 kbps
Test time (minutes)
Loop length (kft) 26AWG
Table A.1-8/TR-100: Margin Verification (400 ft, -140dBm/Hz AWGN) , testprofile
AU_RA_L_30000k
5
Operation in the Presence of Impulse Noise Events
Table A.1-9 describes the testprocedure for the verification of the INP capability of a CO or
CPE modem by testing the operation in the presence of impulse noise events.
Table A.1-9/TR-100: Operation in the Presence of Impulse Noise Events
Test Configuration
Method of Procedure
Expected Result
[1] Connect the CPE to the DSLAM through a loop simulator.
[2] Set the loop simulator to CSA Loop #6 (9000 feet of 26
AWG wire).
[3] Configure
the
DSLAM
with
the
testprofile
AU_RA_I_30000k.
[1] Inject –110 dBm/Hz white noise disturber at the CO end of
the loop.
[2] Train the modems using test AU_RA_I_30000k profile.
[3] Wait for 3 minutes after initialization.
[4] Inject 15 impulses spaced at least 1 second apart into the
circuit at the CO end of the loop simulator over a maximum
of a one minute period. Each impulse shall be a “Burst of
pseudo random AWGN” of 100us duration at a level of 90dBm/Hz differential mode.
[5] Test the CPE modem by repeating the steps above,
injecting the white noise and impulses at the customer end
of the loop.
The number of errored seconds measured after the 3
minute initial period shall be <= 1 for the test to pass.
A.1.3 Verification of downstream bi/gi values
Table A.1-10 describes the testprocedure for the verification of downstream bi/gi values .
© The Broadband Forum. All rights reserved.
53
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.1-10/TR-100: Verification of downstream bi/gi values
Test Configuration
The ATU-C and ATU-R are connected in turn through 3
randomly selected loops (one from each of the sets below)
Testloops:
Set 1 (short loops):
• 1 kft, FEXT – 12 ADSL2Plus, -140 dBm/Hz AWGN
• 2 kft, -140 dBm/Hz AWGN
• 9 kft, NEXT + FEXT - 12 ADSL2Plus, -140 dBm/Hz
AWGN
• 9 kft, NEXT - 5 T1 adjacent, -140 dBm/Hz AWGN
• 9 kft, NEXT – 24HDSL, -140dBm/Hz AWGN
Set 2 (long loops):
• 12 kft, NEXT + FEXT - 12 ADSL2Plus, -140 dBm/Hz
AWGN
• 12 kft, -140 dBm/Hz AWGN
• 13kft, NEXT – 24HDSL, -140dBm/Hz AWGN
• 15 kft, NEXT - 5 T1 adjacent, -140 dBm/Hz AWGN
• 18 kft, -140 dBm/Hz AWGN
Set 3 (bridge tap loops)
• 3 kft with 100 ft BT (24AWG), -140 dBm/Hz AWGN
• 5 kft with 50 ft BT (24AWG), NEXT + FEXT - 12
ADSL2Plus, -140 dBm/Hz AWGN
• 5 kft with 200 ft BT (24AWG), NEXT + FEXT - 12
ADSL2Plus, -140 dBm/Hz AWGN
• 5 kft with 200 ft BT (24AWG), -140 dBm/Hz AWGN
• 9 kft with 100 ft BT (24AWG), -140 dBm/Hz AWGN
•
11 kft with 200 ft BT (24AWG), -140 dBm/Hz
AWGN
[1] Select randomly one testloop from the set 1 above.
[2] Train the modem using the testprofile AU_RA_L_30000k.
[3] From the DSLAM read the bi and gi values requested during
initialization
[4] Repeat steps 1 to 3 for one other randomly selected loop from
set 2.
[5] Repeat steps 1 to 3 for one other randomly selected loop from
set 3.
Expected Result to The RMSGIds shall be calculated in linear scale with gi values as
contained in the BiGi table and then rounded to nearest 100th of a
Pass
dB.
The bi and gi values rounded to the nearest 100th of a dB in the
bits-and-gains table shall comply with the following requirements
(Sections 8.5.2 and 8.6.4 of G.992.3[1])
Method of Procedure
[1] All bi values shall be in the [0 to BIMAXds] (bits) range,
where 8 ≤ BIMAXds ≤ 15.
[2] If bi > 0, then gi shall be in the [–14.54 to +2.50 + EXTGIds]
(dB) range where 0 ≤ EXTGIds ≤ MAXNOMPSDds –
NOMPSDds.
© The Broadband Forum. All rights reserved.
54
ADSL2/ADSL2plus Performance Test Plan
TR-100
[3] RMSGIds ≤ EXTGIds + 1/16th of a dB
[4] If bi > 0, then gi shall be in the [RMSGIds – 2.60 to
RMSGIds + 2.60] (dB) range.
[5] If bi = 0, then gi shall be equal to 0 (linear) or in the [–14.54
to RMSGIds + 0.10] (dB) range.
BIMAXds,
EXTGIds,
MAXNOMPSDds,
NOMPSDds,
RMSGIds are defined in G.992.3[1] Sections 8.5.1 and 8.5.2.
A.1.4 Loop tests with ports set for adaptive rate, automode operation
Tests shall be performed according to the general procedure described in section 7. In
addition to achieving the required rate, both downstream and upstream noise margin values
are to be considered in determining the result of an individual section. It is acknowledged
that achieving a desired noise margin is primarily the responsibility of the receiver. That is,
the DSLAM is primarily responsible for achieving desired upstream noise margins, while
the CPE (modem) is primarily responsible for achieving desired downstream noise
margins. Table A.1-11 outlines the pass/fail criteria on the reported noise margin.
Table A.1-11/TR-100: Noise Margin Chart
Reported Noise Margin
(dB)
<4
>= 4 and < 5
>= 5
>= 6
Requirement
On no test point
On at most 10% of the test
points
On at least 90% of the test
points
On at least 75% of the
downstream test points
All measurements shall be from the DSLAM.
Violation of any of the requirements in the Noise Margin Chart shall constitute a test
section failure. Overall pass/fail criteria for each adaptive rate test section is then as
follows:
• If any reported noise margin is less than 4dB, then the ATU-C/ATU-R pair fails the
noise margin requirements of that section.
• If more than 10% of the reported noise margins are less than 5dB in a section, then
the ATU-C/ATU-R pair fails the noise margin requirements of that section.
• If more than 25% of the reported downstream noise margins are less than 6dB in a
section, then the ATU-C/ATU-R pair fails the noise margin requirements of that
section.
© The Broadband Forum. All rights reserved.
55
ADSL2/ADSL2plus Performance Test Plan
•
•
TR-100
If more than 10% of the net data rates are less than the data rate requirements in a
section, then the ATU-C/ATU-R pair fails the data rate requirements of that
section.
If a testpoint fails on the performance requirement for the following loops:
• Table A.1-14: Loops 13, 14, 15, 16, and 17kft
• Table A.1-15: Loops 13, 14, 15, 16, and 17kft
• Table A.1-16: 15kft Loop with 400ft, 1000ft, and 1500ft Bridged Tap
• Table A.1-17: 13kft Loop
• Table A.1-18: Loops 13, 14, and 15kft
an additional check is done on the sum of measured upstream and downstream net
data rates. If this sum is greater than or equal to the sum of the target upstream and
downstream net data rates, the testpoint is considered a pass. This passing is
indicated in the report as “Testpoint passed due to performance sum passing.”
If the ATU-C/ATU-R pair passes both the data rate and noise margin requirements, it
passes the section; otherwise, it fails the section.
Table A.1-12 lists the number of test points per section corresponding to the 10% and 25%
limits mentioned above.
Table A.1-12/TR-100: Reported Margin Requirements
Section
number
A.1.4.1
A.1.4.2.1
A.1.4.2.2
A.1.4.2.3
A.1.4.3
A.1.4.4
Number of test cases
10% limit
22
44
44
40
20
22
2
4
4
4
2
2
25% limit
(applies to downstream margins only)
3
6
6
5
3
3
A.1.4.1 12 ADSL2plus self NEXT and FEXT noise impairment
Set the noise generator to apply 12 ADSL2plus self NEXT and FEXT and –140dBm/Hz
AWGN noise at both ends of the loop.
For self FEXT only noise, the self FEXT noise shall be injected at the side under test only.
22 individual tests – 20 tests must be passed
© The Broadband Forum. All rights reserved.
56
ADSL2/ADSL2plus Performance Test Plan
TR-100
0
3
0
3
0
5
0
7
0
900
12
ADSL2plus
Dn Fext, 140
dBm/Hz
AWGN
n/a
-140
12
ADSL2plus dBm/Hz
Up Fext, -140 AWGN
dBm/Hz
900
AWGN
-140 dBm/Hz 12
AWGN
ADSL2plus
Dn Fext, 140
dBm/Hz
AWGN
n/a
12
12
ADSL2plus ADSL2plus
Up Fext+Dn Up
Next+Dn
Next, -140
Fext, -140
dBm/Hz
dBm/Hz
AWGN
AWGN
900
12
12
ADSL2plus ADSL2plus
Up Fext+Dn Up
Next+Dn
Next, -140
Fext, -140
dBm/Hz
dBm/Hz
AWGN
AWGN
900
n/a
n/a
n/a
n/a
n/a
n/a
© The Broadband Forum. All rights reserved.
n/a
n/a n/a
n/a
Reported
Operating Mode
Noise Margin,
Reported (dB)
Pass/Fail
n/a n/a
13700
n/a
n/a
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
-140
dBm/Hz
AWGN
Measured
12
ADSL2plus
Up Fext, -140
dBm/Hz
AWGN
-140 dBm/Hz
AWGN
testprofile AU_RA_I_30000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate
(kbps)
Expected
Noise Applied on Remote
side
1
0
Noise Applied on CO side
1
Bridge tap length 24 AWG
(ft)
Loop length 26 AWG (kft)
Table A.1-13/TR-100: ADSL2plus self-next and fext and -140dBm/Hz AWGN noise
12254
11001
9284
57
12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
5
50 12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
5
100 12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
5
150 12
ADSL2plus
Up Fext+Dn
Next, -140
800
5712
800
3576
913
10796
900
10947
900
11001
© The Broadband Forum. All rights reserved.
Reported
Operating Mode
0
6969
Noise Margin,
Reported (dB)
12
888
Pass/Fail
12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
Measured
0
Expected
10
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Noise Margin,
Reported (dB)
12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
Pass/Fail
0
Measured
Noise Applied on Remote
side
9
TR-100
testprofile AU_RA_I_30000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate
(kbps)
Expected
Noise Applied on CO side
Bridge tap length 24 AWG
(ft)
Loop length 26 AWG (kft)
ADSL2/ADSL2plus Performance Test Plan
58
12
ADSL2plus
Up Fext+Dn
Next, -140
dBm/Hz
AWGN
Reported
Operating Mode
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Pass/Fail
Expected
Expected
Fext, -140
dBm/Hz
AWGN
12
ADSL2plus
Up
Next+Dn
Fext, -140
dBm/Hz
AWGN
900
Noise Margin,
Reported (dB)
dBm/Hz
AWGN
Measured
Noise Applied on Remote
side
200
TR-100
testprofile AU_RA_I_30000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate
(kbps)
Noise Applied on CO side
5
Bridge tap length 24 AWG
(ft)
Loop length 26 AWG (kft)
ADSL2/ADSL2plus Performance Test Plan
10790
A.1.4.2 AWGN –140 dBm/Hz noise impairment
Set the noise generator to apply –140dBm/Hz AWGN noise at both ends of the loop.
A.1.4.2.1
26AWG loop with testprofile AU_RA_L_30000k
44 individual tests – 40 tests must be passed
0.1
1000
23412
0.4
1000
23768
1
1000
23616
2
1000
22740
3
1000
21551
4
1000
20287
© The Broadband Forum. All rights reserved.
Reported
Operating Mode
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
testprofile AU_RA_L_30000
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop length 26 AWG
(kft)
Table A.1-14/TR-100: -140dBm/Hz AWGN noise, 26awg loop, testprofile
AU_RA_L_30000
59
A.1.4.2.2
1000
18186
6
1000
16456
7
1000
14368
8
1000
11236
9
994
9080
10
976
6734
11
954
5521
12
900
4167
13
846
3264
14
779
2717
15
516
1952
16
430
1387
17
344
799
18
328
464
19
239
235
20
150
212
Reported
Operating Mode
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
5
Noise Margin,
Reported (dB)
TR-100
testprofile AU_RA_L_30000
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop length 26 AWG
(kft)
ADSL2/ADSL2plus Performance Test Plan
26AWG loop with testprofile AU_RA_I_30000k
44 individual tests – 40 tests must be passed
© The Broadband Forum. All rights reserved.
60
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.1-15/TR-100: -140dBm/Hz AWGN noise, 26awg loop, testprofile
AU_RA_I_30000k
Expected
0.1
900
18526
0.4
900
18732
1
900
18314
2
900
18200
3
900
17123
4
900
16392
5
900
15613
6
900
14095
7
900
13300
8
900
11308
9
850
9188
10
800
6704
11
750
5364
12
700
4496
13
650
3496
14
600
2292
15
550
2022
16
500
1318
17
376
464
18
356
355
19
336
246
20
227
178
A.1.4.2.3
Pass/Fail
Measured
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Downstream
Noise Margin,
Reported (dB)
Upstream
Reported
Operating Mode
Loop length 26 AWG
(kft)
testprofile AU_RA_I_30000k
26AWG loop with 24AWG bridge taps and testprofile AU_RA_L_30000k
40 individual tests – 36 tests must be passed
© The Broadband Forum. All rights reserved.
61
ADSL2/ADSL2plus Performance Test Plan
TR-100
100
1000
21971
3
100
1000
19423
5
100
1000
16631
7
100
1000
12862
9
100
900
8125
11
100
900
5270
3
200
1000
20932
5
200
1000
17812
7
200
1000
13156
9
200
1000
8221
11
200
893
4179
12
250
803
3293
12
750
763
3842
12
1500
492
3460
15
400
516
1252
15
1000
491
1469
15
1500
336
1748
17.5
150
336
428
17.5
800
240
524
17.5
1500
24
508
Pass/Fail
Measured
Expected
Pass/Fail
Measured
Expected
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Bridge tap length 24
AWG (ft)
0.5
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Loop length 26 AWG
(kft)
testprofile AU_RA_L_30000k
Upstream
Downstream
Reported
Operating Mode
Table A.1-16/TR-100: -140dBm/Hz AWGN noise, 26awg loop with 24awg bridge taps
A.1.4.3 24 HDSL NEXT impairment
Set the noise generator to apply 24 HDSL NEXT impairment and –140dBm/Hz AWGN
noise at both ends of the loop.
20 individual tests – 18 tests must be passed
© The Broadband Forum. All rights reserved.
62
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.1-17/TR-100: 24HDSL NEXT and -140dBm/Hz AWGN noise
Bridge tap length 24
AWG (ft)
Expected
1
0
1000
20000
3
0
1000
19484
5
0
879
17506
7
0
588
12988
9
0
560
7184
11
0
366
3619
13
0
192
1624
5
100
867
15443
5
200
807
17173
6
500
665
14443
Pass/Fail
Measured
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Downstream
Noise Margin,
Reported (dB)
Upstream
Reported
Operating Mode
Loop length 26 AWG
(kft)
testprofile AU_RA_L_30000
A.1.4.4 5 T1 NEXT adjacent binder impairment
Set the noise generator to apply 5 T1 Next(adjacent binder) impairment and –140dBm/Hz
AWGN noise at both ends of the loop.
22 individual tests – 20 tests must be passed
© The Broadband Forum. All rights reserved.
63
ADSL2/ADSL2plus Performance Test Plan
TR-100
Loop length 26 AWG
(kft)
Bridge tap length 24
AWG (ft)
Expected
Table A.1-18/TR-100: 5 T1(adj) NEXT and -140dBm/Hz AWGN noise
1
0
1000
20321
3
0
1000
16235
5
0
1000
9648
7
0
1000
4702
9
0
994
2626
13
0
846
796
14
0
620
515
15
0
456
86
5
100
1000
8243
5
200
1000
10082
6
500
1000
7447
Pass/Fail
Measured
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Downstream
Noise Margin,
Reported (dB)
Upstream
Reported
Operating Mode
Testprofile AU_RA_L_30000k
A.1.5 Loop tests with ports set for adaptive rate, single operating mode
Tests shall be performed according to the general procedure described in section 7.
CO and CPE ends shall operate in singlemode over POTS using the FDM spectral mask.
The specific testprofiles are changed to enable only the operating mode indicated in each
table for that specific test.
In addition to achieving the required rate, both downstream and upstream noise margin
values are to be considered in determining the result of an individual section. It is
acknowledged that achieving a desired noise margin is primarily the responsibility of the
receiver. That is, the DSLAM is primarily responsible for achieving desired upstream noise
margins, while the CPE (modem) is primarily responsible for achieving desired
downstream noise margins. Table A.1-11 outlines the reported noise margin pass/fail
criteria.
All measurements shall be from the DSLAM.
Violation of any of the requirements in the Noise Margin Chart shall constitute a test
section failure. Overall pass/fail criteria for each adaptive rate test section is then as
follows:
• If any reported noise margin is less than 4dB, then the ATU-C/ATU-R pair fails the
noise margin requirements of that section.
© The Broadband Forum. All rights reserved.
64
ADSL2/ADSL2plus Performance Test Plan
TR-100
•
If more than 10% of the reported noise margins are less than 5dB in a section, then
the ATU-C/ATU-R pair fails the noise margin requirements of that section.
• If more than 25% of the reported downstream noise margins are less than 6dB in a
section, then the ATU-C/ATU-R pair fails the noise margin requirements of that
section.
• If more than 10% of the net data rates are less than the data rate requirements in a
section, then the ATU-C/ATU-R pair fails the data rate requirements of that
section.
If the ATU-C/ATU-R pair passes both the data rate and noise margin requirements, it
passes the section; otherwise, it fails the section.
Table A.1-19 lists the number of test points per section corresponding to the 10% and 25%
limits mentioned above.
Table A.1-19/TR-100: Reported Margin Requirements (single operating mode)
Section
number
A.1.5.1
A.1.5.2
A.1.5.3
A.1.5.4
Number of test cases
10% limit
8
16
6
6
1
2
1
1
25% limit
(applies to downstream
margins only)
1
2
1
1
A.1.5.1 12 Self NEXT and FEXT noise impairment
8 individual tests – 7 tests must be passed
© The Broadband Forum. All rights reserved.
65
ADSL2/ADSL2plus Performance Test Plan
TR-100
1
1
5
12
15
testprofile AU_RA_I_30000k
12
ADSL2plus
Up Fext, - -140 dBm/Hz
AWGN
140
dBm/Hz
AWGN
12
ADSL2plus
-140
dBm/Hz
Dn Fext, -140
AWGN
dBm/Hz
AWGN
12
12
ADSL2plus
ADSL2plus
Up
Up Next+Dn
Fext+Dn
Fext, -140
Next, -140
dBm/Hz
dBm/Hz
AWGN
AWGN
12
12
ADSL2plus
ADSL2plus
Up
Up Next+Dn
Fext+Dn
Fext, -140
Next, -140
dBm/Hz
dBm/Hz
AWGN
AWGN
12 ADSL2
12 ADSL2
Annex-L
Annex-L Up
Up
Next+Dn
Fext+Dn
Fext, -140
Next, -140
dBm/Hz
dBm/Hz
AWGN
AWGN
900
n/a
n/a
n/a
n/a
n/a
n/a
Operating Mode
Noise Margin,
Reported (dB)
Pass/Fail
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Expected
Sync Rate
(kbps)
Downstream
Noise Margin,
Reported (dB)
Upstream
Measured
Noise Applied on Remote
side
Noise Applied on CO side
Loop length 26 AWG (kft)
Table A.1-20/TR-100: 12 Self NEXT and FEXT and –140dbm/Hz AWGN noise
n/a ADSL2plus
n/a 13700
ADSL2plus
900
11001
ADSL2plus
800
3576
ADSL2
1348
ADSL2
AnnexL
(wide
US
PSD)
464
A.1.5.2 AWGN –140 dBm/Hz noise impairment
Set the noise generator to apply –140dBm/Hz AWGN noise at both ends of the loop.
© The Broadband Forum. All rights reserved.
66
ADSL2/ADSL2plus Performance Test Plan
TR-100
16 individual tests – 14 tests must be passed
Table A.1-21/TR-100: -140dBm/Hz AWGN noise, testprofile AU_RA_L_30000k
Bridge tap length 24
AWG (ft)
Expected
1
0
1000
23616
ADSL2plus
5
0
1000
18186
ADSL2plus
12
0
846
4167
ADSL2
524
ADSL2
AnnexL
(wide US
PSD)
17.5
800
Downstream
240
Pass/Fail
Expected
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Upstream
Operating Mode
Loop length 26 AWG
(kft)
Testprofile AU_RA_L_30000k
Pass/Fail
Measured
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Downstream
Noise Margin,
Reported (dB)
Upstream
Operating Mode
testprofile AU_RA_I_30000k
Expected
Loop length 26 AWG
(ft)
Table A.1-22/TR-100: -140dBm/Hz AWGN noise, testprofile AU_RA_I_30000k
1
900
18314
ADSL2plus
5
900
15613
ADSL2plus
12
700
4496
ADSL2
355
ADSL2
AnnexL
(wide US
PSD)
18
356
© The Broadband Forum. All rights reserved.
67
ADSL2/ADSL2plus Performance Test Plan
TR-100
A.1.5.3 24 HDSL NEXT impairment
Set the noise generator to apply 24 HDSL NEXT impairment and –140dBm/Hz AWGN
noise at both ends of the loop.
6 individual tests – 5 tests must be passed
Operating Mode
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
testprofile AU_RA_L_30000
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop length 26 AWG
(kft)
Table A.1-23/TR-100: 24HDSL NEXT and -140dBm/Hz AWGN noise
1
1000
20000
ADSL2plus
5
879
17506
ADSL2plus
13
192
1624
ADSL2
A.1.5.4 5 T1 NEXT adjacent binder impairment
Set the noise generator to apply 5 T1 Next(adjacent binder) impairment and –140dBm/Hz
AWGN noise at both ends of the loop.
6 individual tests – 5 tests must be passed
© The Broadband Forum. All rights reserved.
68
ADSL2/ADSL2plus Performance Test Plan
TR-100
Loop length 26 AWG
(kft)
Expected
Table A.1-24/TR-100: 5 T1(adj) NEXT and -140dBm/Hz AWGN noise
1
1012
20321
ADSL2plus
8
922
3953
ADSL2
15
456
86
Pass/Fail
Measured
Sync Rate (kbps)
Expected
Pass/Fail
Measured
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Downstream
Noise Margin,
Reported (dB)
Upstream
Operating Mode
testprofile AU_RA_L_30000
ADSL2
AnnexL
(wide US
PSD)
A.1.6 Loop Tests with Ports Set For Fixed Rate
This section is for further study.
A.1.7 ADSL2plus Coexistence with Phoneline Networking
The test case described in table A.1-25 is intended to verify compatibility when
ADSL2plus and a phoneline networking system as defined in G.pnt/G.9954[14], when both
use the same inside wire pair
The test is to be performed on three different loop length of 26AWG pairs: 3kft, 5kft, and
7kft.
The pass/fail criterion is the bit error ratio (BER) at the physical layer. CRC error counts
are used to estimate BER at the physical layer.
The injected G.pnt noise shall have the PSD as defined in G.9954 [14] spectral Mask #3.
The PSD breakpoints are defined in [14] Table 6-14/G.9954 and is shown in [14] Figure 619/G.9954
SRA (seamless rate adaptation) capability shall be switched off during the test.
© The Broadband Forum. All rights reserved.
69
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.1-25/TR-100 : ADSL2plus and G.pnt coexistence
Test
The test set up is as shown in figure 14.
Configuration
Details of Test [1] Configure CPE and DSLAM according to the testprofile
AU_RA_I_30000k.
Procedure
[2] Configure a testloop of 3kft 26AWG. Connect CPE and DSLAM to the
testloop. Apply the G.pnt noise at the ATU-R,
[3] Force initialization and wait for modem to sync.
[4] Wait for 3 minutes after initialization for bitswaps to settle.
[5] Record the net data rate R(kbps).
[6] Switch the noise to -140dBm/Hz AWGN and wait for 1 minute.
[7] Record the net data rate R1(kbps) and count the number of CRC errors
(C1) in the period of T, calculated using the following formula:
Expected
Result
T = 400/(R*1000*1e-7*60) (min)
and estimate the BER (P1) using Table7-12, interleaved path.
[8] Switch the noise back to G.pnt and wait for 1 minute.
[9] Record the net data rate R2(kbps) and count the number of CRC errors
(C2) in the period of T and estimate the BER (P2) using Table 7-12,
interleaved path.
[10] Repeat steps [2] to [9] for testloops 5kft 26Awg and 7 kft 26AWG
The CPE modem passes the G.pnt co-existence test, if for each testloop,
(a) The modem does not re-initialize during the test
(b) The estimated BER P1 and P2 are < 1.5e-7
(c) R = R1 = R2
© The Broadband Forum. All rights reserved.
70
ADSL2/ADSL2plus Performance Test Plan
TR-100
Figure 14/TR-100: G.pnt test setup
A.1.8 REIN noise impairment
This section is for further study.
© The Broadband Forum. All rights reserved.
71
ADSL2/ADSL2plus Performance Test Plan
A.2
TR-100
European Test set
The specific testprofiles as defined in table 7.3 shall be used.
The performance on short loops is based on the ADSL-like (G.992.1) power cutback
mechanism.
Two performance classes based on the support of optional extended framing parameters for
the use of high delay profile with INP = 2 are defined.
Class A: The DSLAM – modem combination does support the optional extended
framing parameters acc. G.992.3[1] and G.992.5[2]
Class B: The DSLAM – modem combination does not support the optional
extended framing parameters acc. G.992.3[1] and G.992.5[2]
The performance tests have to be performed once and after filling in the results table the
DSLAM – modem combination is allocated to the one or the other performance class.
A.2.1 Noise and noise injection
Noise shall be injected as specified in ETSI TS101388[8], with simultaneous noise
injection at both ends of the loop.
The noise injection shall be calibrated as defined in ETSI TS101388[8].
The simultaneous used generators shall be G1 (NEXT noise), G2 (FEXT noise), G4 (white
noise).
For white noise set generator G4 at both ends of the total loop at –140 dBm/Hz.
A.2.1.1 Noise Generator settings G1, G2 and G4 for ADSL2 noise FA or FD
impairment,
All measurements with noise model FA and FD shall be done with the noise model
calculated for the corresponding length of the ETSI loop selected as indicated in the
applicable testsection. Set noise generator G4 to apply a white noise disturber at both ends
of the loop at -140 dBm/Hz.
Set noise generator G1 and G2 for ETSI FA noise:
At CO side:
• G1 = X.LT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 12,
with NEXT coupling function.
• G2 = X.NT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 13,
with FEXT coupling function.
At CPE side:
• G1 = X.NT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 13,
with NEXT coupling function.
• G2 = X.LT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 12,
with FEXT coupling function.
A.2.1.2 Noise Generator settings G1, G2 and G4 for ADSL2 noise FB impairment,
All measurements with noise model FB shall be done with the noise model FB calculated
for 3000 m ETSI loop #1.
© The Broadband Forum. All rights reserved.
72
ADSL2/ADSL2plus Performance Test Plan
TR-100
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
Set noise generator G1 and G2 for ETSI FB noise:
At CO side:
• G1 = X.LT.FB as defined in TS 101 388 [8], section. 5.3.4.1.3, table 12, with
NEXT coupling function.
• G2 = X.NT.FB as defined in TS 101 388 [8], section. 5.3.4.1.3, table 13, with
FEXT coupling function.
At CPE side:
• G1 = X.NT.FB as defined in TS 101 388 [8], section. 5.3.4.1.3, table 13, with
NEXT coupling function.
• G2 = X.LT.FB as defined in TS 101 388 [8], section. 5.3.4.1.3, table 12, with
FEXT coupling function.
A.2.1.3 Noise Generator settings G1, G2 and G4 for ADSL2plus noise FA, FB and FD
impairment,
The ADSL2plus noise FA, FB and FD shall be calculated using the “Definition of the
cross-talk coupling functions” from TS101388[8] section 5.3.2 table 5.
The template PSD masks for ADSL2plus noise FA, FB and FD shall be as defined in annex
D table D.1-1 for NT side and table D.1-2 for LT side.
Set noise generator G4 to apply white noise disturber at both ends of the total loop at -140
dBm/Hz.
All measurements with noise model FA, FB and FD shall be done with the noise model
calculated for the corresponding length of ETSI loop #1.
Set noise generator G1 and G2 for ETSI FA/FB/FC ADSL2plus noise:
At CO side:
• G1 = X.LT. FA/FB/FD for ADSL2plus, with NEXT coupling function.
• G2 = X.NT. FA/FB/FD for ADSL2plus, with FEXT coupling function.
At CPE side:
• G1 = X.NT. FA/FB/FD for ADSL2plus, with NEXT coupling function.
• G2 = X.LT. FA/FB/FD for ADSL2plus, with FEXT coupling function.
A.2.2 CPE Margin verification tests
Section 7.5 gives a description and testprocedure for the margin verification tests in tables
A.2-1 up to A.2-7
A.2.2.1 Noise FB ADSL2 impairments at 750 m.
Noise FB settings as defined in section A.2.1.2, using testprofile A2_RA_F_16000k
© The Broadband Forum. All rights reserved.
73
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-1/TR-100: Noise FB ADSL2 impairment, testprofile A2_RA_F_16000k, at
750 m
Test time
Loop
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 8932
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
750
Estimated BER,
based on
equations in Table
7-12, fast path
5
A.2.2.2 Noise FB ADSL2 impairments at 2750 m.
Noise FB settings as defined in section A.2.1.2, using testprofile A2_RA_I1/2_16000k
Table A.2-2/TR-100: Noise FB ADSL2 impairment, testprofile A2_RA_I1/2_16000k,
at 2750 m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 2156
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Achieved DS
net data rate
(test end)
2750
Estimated BER,
based on
equations in Table
7-12, interleaved
path
Pass
/
Fail
If:
Estimated_BER <
1. 5e-7
then PASS
else FAIL
35
A.2.2.3 ADSL2 Random selected loop (Optional test)
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2 as defined in sections A2.1.1, using testprofile A2_RA_I1/2_16000k.
© The Broadband Forum. All rights reserved.
74
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-3/TR-100: Selected Test Loop, testprofile A2_RA_I1/2_16000k
Test time
ETSI
(minutes)
Loop
Selected Calculate
based on
connect
rate.
Achieved
DS net data
rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
DS net data
rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Loop
Selected
Note: The table for section A.2-3 is an example table. The specific test loop
selected and noise shape used should be indicated in the table. The connect rate
and also test time calculated should be indicated in the table A.2-3.
A.2.2.4 Noise FD ADSL2 impairments at 5250m (optional test).
Noise FD settings as defined in section A.2.1.1, using test profile A2L_RA_I_16000k
Table A.2-4/TR-100: Noise FD ADSL2 impairment, test profile A2L_RA_I_16000k, at
5250 m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 448 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test end)
5250
Estimated BER,
based on
equations in Table
7-12, interleaved
path
Pass
/
Fail
If:
Estimated_BER <
1. 85e-7
then PASS
else FAIL
60
NOTE: Due to the low, achievable DS data rates under this loop and noise
condition, the observation of 10 error events would require approximately 150
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of observed
error events is reduced to 4. To remain consistent with previous confidence
© The Broadband Forum. All rights reserved.
75
ADSL2/ADSL2plus Performance Test Plan
TR-100
levels of estimated BER, the range of allowed estimated BER is increased from
1.5-e7 to 1.85e-7
A.2.2.5 Noise FB ADSL2plus impairments at 250 m.
Noise FB settings as defined in section A.2.1.3, using testprofile A2P_RA_F_30000k
Table A.2-5/TR-100: Noise FB ADSL2plus impairment, testprofile
A2P_RA_F_30000k, at 250 m
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 14648
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Achieved DS
net data rate
(test end)
250
Estimated BER,
based on
equations in Table
7-12, fast path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
5
A.2.2.6 Noise FB ADSL2plus impairments at 1750 m.
Noise FB settings as defined in section A.2.1.3, using testprofile A2P_RA_I_30000k
© The Broadband Forum. All rights reserved.
76
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-6/TR-100: Noise FB ADSL2plus impairment, testprofile
A2P_RA_I_30000k, at 1750 m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 9296
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Pass
/
Fail
If:
Estimated_BER <
1. 5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
1750
Estimated BER,
based on
equations in Table
7-12, interleaved
path
10
A.2.2.7 ADSL2plus Random selected loop (Optional test)
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2plus as defined is section A.2.1.3, using testprofile A2P_RA_I_30000k
Table A.2-7/TR-100: Margin Tests for Selected ETSI Loop #3 – 7, testprofile
A2P_RA_I_30000k
Test time
ETSI
(minutes)
Loop
Selected Calculate
based on
connect
rate.
Achieved
DS net data
rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
DS net data
rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Loop
Selected
Note: The table is an example table. The specific test loop selected and noise
shape used shall be indicated in table A.2-7. The connect rate and also test time
calculated shall also be indicated in table A.2-7
© The Broadband Forum. All rights reserved.
77
ADSL2/ADSL2plus Performance Test Plan
TR-100
A.2.3 CO Margin Verification (Optional)
Section 7.5 gives a description and testprocedure for the margin verification tests in tables
A.2-8 up to A.2-11.
A.2.3.1 Noise FB ADSL2 impairment at 750m.
Noise FB settings as defined in section A.2.1.2, using test profile A2_RA_F_16000k
Table A.2-8/TR-100: Upstream margin test with noise FB ADSL2 impairment, test
profile A2_RA_F_16000k, at 750 meters
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 1048 kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test end)
750
Estimated BER,
based on
equations in Table
7-12, fast path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
25
A.2.3.2 Noise FB ADSL2 impairment at 2750m.
Noise FB settings as defined in section A.2.1.2, using test profile A2_RA_I1/2_16000k
© The Broadband Forum. All rights reserved.
78
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-9/TR-100: Upstream margin tests with noise FB ADSL2 impairment, test
profile A2_RA_I1/2_16000k, at 2750 meters
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 536 kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test end)
2750
Estimated BER,
based on
equations in Table
7-12, interleaved
path
Pass
/
Fail
If:
Estimated_BER <
1.75e-7
then PASS
else FAIL
60
NOTE: Due to the low achievable US data rates under this loop and noise
condition, the observation of 10 error events would require approximately 85
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of observed
error events is reduced to 5. To remain consistent with previous confidence
levels of estimated BER, the range of allowed estimated BER is increased from
1.5-e7 to 1.75e-7
A.2.3.3 Noise FB ADSL2plus impairment at 250m.
Noise FB Settings as defined in section A.2.1.3, using test profile A2P_RA_F_30000k
© The Broadband Forum. All rights reserved.
79
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-10/TR-100: Upstream margin test with noise FB ADSL2plus impairment,
test profile A2P_RA_F_30000k, at 250 meters
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 1048
kbps
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test start)
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved US
net data rate
(test end)
250
Estimated BER,
based on
equations in Table
7-12, fast path
25
A.2.3.4 Noise FB ADSL2plus impairment at 1750m.
FB noise settings as defined in section A.2.1.3, using test profile A2P_RA_I_30000k
Table A.2-11/TR-100: Upstream margin tests with noise FB ADSL2plus impairment
at 1750 meters, test profile A2P_RA_I_30000k
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 804 kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test end)
1750
Estimated BER,
based on
equations in Table
7-12, interleaved
path
Pass
/
Fail
If:
Estimated_BER <
1.75e-7
then PASS
else FAIL
45
NOTE: Due to the low, achievable DS data rates under this loop and noise
condition, the observation of 10 error events would require approximately 85
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of observed
error events is reduced to 5. To remain consistent with previous confidence
© The Broadband Forum. All rights reserved.
80
ADSL2/ADSL2plus Performance Test Plan
TR-100
levels of estimated BER, the range of allowed estimated BER is increased from
1.5-e7 to 1.75e-7
A.2.4 Verification of downstream bi/gi values
A.2.4.1 ADSL2
Table A.2-12 describes the testprocedure for the verification of downstream bi/gi values in
ADSL2 operating mode.
Table A.2-12/TR-100: Verification of downstream bi/gi values for ADSL2
The ATU-C and ATU-R are connected in turn through the following
loops:
[1] ETSI loop #1, 1250 m, -140 dBm/Hz AWGN, testprofile
A2_RA_F_16000k
[2] ETSI loop #1, 3250 m, -140 dBm/Hz AWGN, testprofile
A2_RA_I/2_16000k
[3] ETSI loop #1, 750 m, noise FB, testprofile A2_RA_I_16000k
[4] ETSI loop #1, 2750 m, noise FB, testprofile A2_RA_F_16000k
Method
of [1] Select test loop 1 from the set above.
[2] Train the modem in the test profile defined for the selected loop.
Procedure
[3] From the DSLAM read the bi and gi values requested during
initialization
[4] Repeat steps 1 to 3 for other test loops in the set.
Expected Results The RMSGIds shall be calculated in linear scale with gi values as
contained in the BiGi table and then rounded to nearest 100th of a dB.
The bi and gi values rounded to the nearest 100th of a dB in the bits-andgains table shall comply with the following requirements (Sections 8.5.2
and 8.6.4 of G.992.3[1])
Test
Configuration
[1] All bi values shall be in the [0 to BIMAXds] (bits) range, where 8 ≤
BIMAXds ≤ 15.
[2] If bi > 0, then gi shall be in the [–14.54 to +2.50 + EXTGIds] (dB)
range where 0 ≤ EXTGIds ≤ MAXNOMPSDds – NOMPSDds.
[3] RMSGIds ≤ EXTGIds + 1/16th of a dB.
[4] If bi > 0, then gi shall be in the [RMSGIds – 2.60 to RMSGIds +
2.60] (dB) range.
[5] If bi = 0, then gi shall be equal to 0 (linear) or in the [–14.54 to
RMSGIds + 0.10] (dB) range.
BIMAXds, EXTGIds, MAXNOMPSDds, NOMPSDds, RMSGIds are
defined in G.992.3[1] Sections 8.5.1 and 8.5.2.
A.2.4.2 ADSL2plus
Table A.2-13 describes the testprocedure for the verification of downstream bi/gi values
in ADSL2plus operating mode.
© The Broadband Forum. All rights reserved.
81
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-13/TR-100: Verification of downstream bi/gi values for ADSL2plus
Test
Configuration
Method
Procedure
The ATU-C and ATU-R are connected in turn through the following
loops:
[1] ETSI loop #1, 250 m, -140 dBm/Hz AWGN, testprofile
A2P_RA_F_30000k
[2] ETSI loop #1, 1750 m, -140 dBm/Hz AWGN, testprofile
A2P_RA_I_30000k
[3] ETSI loop #1, 250 m, noise FB, testprofile A2P_RA_I_30000k
[4] ETSI loop #1, 1750 m, noise FB, testprofile A2P_RA_F_30000k
of [1] Select test loop 1 from the set above.
[2] Train the modem in the test profile defined for the selected loop.
[3] From the DSLAM read the bi and gi values requested during
initialization
[4] Repeat steps 1 to 3 for other test loops in the set.
Expected Results
The RMSGIds shall be calculated in linear scale with gi values as
contained in the BiGi table and then rounded to nearest 100th of a dB.
The bi and gi values rounded to the nearest 100th of a dB in the bits-andgains table shall comply with the following requirements (Sections 8.5.2
and 8.6.4 of G.992.3[1])
[1] All bi values shall be in the [0 to BIMAXds] (bits) range, where 8 ≤
BIMAXds ≤ 15.
[2] If bi > 0, then gi shall be in the [–14.54 to +2.50 + EXTGIds] (dB)
range where 0 ≤ EXTGIds ≤ MAXNOMPSDds – NOMPSDds.
[3] RMSGIds ≤ EXTGIds + 1/16th of a dB
[4] If bi > 0, then gi shall be in the [RMSGIds – 2.60 to RMSGIds +
2.60] (dB) range.
[5] If bi = 0, then gi shall be equal to 0 (linear) or in the [–14.54 to
RMSGIds + 0.10] (dB) range.
BIMAXds, EXTGIds, MAXNOMPSDds, NOMPSDds, RMSGIds are
defined in G.992.3[1] Sections 8.5.1 and 8.5.2.
A.2.5 Loop Tests with Ports Set for Adaptive Rate
The same pass/fail criteria as defined in section A.1-4 apply.
Table A.2-14 lists the number of test points per section corresponding to the 10% and
25% reported margin requirements limits mentioned above.
© The Broadband Forum. All rights reserved.
82
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-14/TR-100: Reported Margin Requirements
Section
number
Number of test cases
10% limit
32
28
32
36 (class A)
28 (class B)
A.2.5.1
A.2.5.2
A.2.5.3
A.2.5.4
3
3
3
4 (class A)
3 (class B)
25% limit
(applies to downstream margins only)
4
4
4
5 (class A)
4 (class B)
A.2.5.1 ADSL2 white noise impairment
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
16 individual tests – 14 tests must be passed
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
A2_RA_F_16000k
Downstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
1048
1048
1048
1048
1004
800
576
378
Pass/Fail
250
1250
1750
2500
3250
4000
4750
5500
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
Table A.2-15/TR-100: White noise testprofile A2_RA_F_16000k
12232
12232
11952
9968
6280
3292
1688
604
16 individual tests – 14 tests must be passed
© The Broadband Forum. All rights reserved.
83
ADSL2/ADSL2plus Performance Test Plan
TR-100
250
1250
1750
2500
3250
4000
4750
5500
1048
1048
1048
1048
1004
820
592
384
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2_RA_I1/2_16000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
Table A.2-16/TR-100: White noise, testprofile A2_RA_I1/2_16000k
12184
12184
11892
10276
6700
3576
1868
760
A.2.5.2 ADSL2 noise FB impaiment
FB noise settings as defined in section A.2.1.2
Note: Short loop power cutback in ADSL2 can be determined on request by
either side (DSLAM and CPE), depending on the receiver front-end
capabilities. I.e. a short loop power cutback requested by CPE determines
downstream aggregate transmit power, which mainly impacts downstream
performance. Downstream performance targets defined in Table A.2-17 and
A.2-18 take into account a maximum power cutback of 12 dB.
14 individual tests – 12 tests must be passed
© The Broadband Forum. All rights reserved.
84
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-17/TR-100: Noise FB impairment, testprofile A2_RA_F_16000k
Noise Margin,
Reported (dB)
Pass/Fail
Pass/Fail
Measured
Measured
1048
1048
964
828
672
504
340
Downstream
Sync Rate (kbps)
Expected
Expected
250
750
1250
1750
2250
2750
3250
Noise Margin,
Reported (dB)
Loop Length
(m, ETSI loop #1)
A2_RA_F_16000k
Upstream
Sync Rate (kbps)
7960
8932
8096
6592
4272
1596
64
14 individual tests – 12 tests must be passed
Table A.2-18/TR-100: Noise FB impairment, testprofile A2_RA_I1/2_16000k
A2_RA_I1/2_16000k
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
1048
1048
972
836
696
536
376
Downstream
Sync Rate (kbps)
Pass/Fail
Expected
250
750
1250
1750
2250
2750
3250
Measured
Loop Length
(m, ETSI loop #1)
Upstream
Sync Rate (kbps)
8052
9244
8348
6912
4616
2156
96
A.2.5.3 ADSL2plus white noise impairment
Set noise generator G4 to apply white noise disturber at both ends of the total loop at -140
dBm/Hz.
16 individual tests – 14 tests must be passed
© The Broadband Forum. All rights reserved.
85
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-19/TR-100: white noise impairment, testprofile A2P_RA_F_30000k
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Downstream
Sync Rate (kbps)
Expected
Expected
Noise Margin,
Reported (dB)
1048
1048
1048
1048
1004
800
576
378
Pass/Fail
250
1250
1750
2500
3250
4000
4750
5500
Measured
Loop Length
(m, ETSI loop #1)
A2P_RA_F_30000k
Upstream
Sync Rate (kbps)
23800
22048
17340
10432
6100
3292
1688
604
16 individual tests – 14 tests must be passed (applies to Class A implementations only)
1750
944
16000
2500
944
10100
3250
944
6100
4000
820
3200
4750
592
1800
5500
384
700
© The Broadband Forum. All rights reserved.
Pass/Fail
18800
Measured
944
Expected
1250
Noise Margin,
Reported (dB)
20600
Pass/Fail
944
Measured
250
Expected
Loop Length
(m, ETSI loop #1)
A2P_RA_I_30000k
Downstream
Sync Rate (kbps)
Upstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Table A.2-20/TR-100: white noise impairment, testprofile A2P_RA_I_30000k
86
ADSL2/ADSL2plus Performance Test Plan
TR-100
16 individual tests – 14 tests must be passed (optional for class A implementations)
250
1250
1750
2500
3250
4000
4750
5500
1048
1048
1048
1048
1004
820
592
384
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_RA_I1/2_30000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
Table A.2-21/TR-100: white noise impairment, testprofile A2P_RA_I1/2_30000k
23296
21984
17444
10964
6480
3576
1868
760
A.2.5.4 ADSL2plus FB noise impairment
FB noise settings as defined in section A.2.1.3
14 individual tests – 12 tests must be passed
© The Broadband Forum. All rights reserved.
87
ADSL2/ADSL2plus Performance Test Plan
TR-100
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
A2P_RA_F_30000k
Downstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
1048
1048
1048
910
776
560
308
Pass/Fail
100
250
750
1250
1750
2500
3250
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
Table A.2-22/TR-100: Noise FB ADSL2plus impairment testprofile
A2P_RA_F_30000k
16084
14648
13548
11552
8600
2820
78
14 individual tests – 12 tests must be passed (applies to Class A implementations only)
14296
750
944
13600
1250
928
11800
1750
804
9296
2500
588
3330
3250
348
200
Pass/Fail
944
Measured
250
Expected
14296
Noise Margin,
Reported (dB)
944
Pass/Fail
100
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
A2P_RA_I_30000k
Downstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Table A.2-23/TR-100: Noise FB ADSL2plus impairment, testprofile
A2P_RA_I_30000k
14 individual tests – 12 tests must be passed
© The Broadband Forum. All rights reserved.
88
ADSL2/ADSL2plus Performance Test Plan
TR-100
If the last six test points are not performed, as being optional, the tests to be passed shall
be: 8 individual tests – 7 tests must be passed
100
250
750
1250
1750*
2500*
3250*
1048
1048
1048
928
804
588
348
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
A2P_RA_I1/2_30000k
Downstream
Sync Rate (kbps)
Pass/Fail
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #1)
Table A.2-24/TR-100: Noise FB ADSL2plus impairment, testprofile
A2P_RA_I1/2_30000k
16640
15476
14304
12444
9488
3330
200
* : optional testpoints for class A implementations
A.2.6 Loop Tests with Ports Set for Fixed Rate
Pass/fail criteria in this section:
• If upstream or downstream reported noise margin on any test point is >= 4dB and
< 6dB, then this test point shall be re-tested 3 times. If a re-test is performed, then
the maximum reported downstream noise margin achieved during testing shall be
recorded as reported noise margin for this test point, along with the associated
upstream noise margin provided that this is >=6dB.
• If the final upstream or downstream reported noise margin (from step 1 above) is
less than 6dB on any test point, then the ATU-C/ATU-R pair fails that section.
A.2.6.1 ADSL2 FB noise impairment
The ADSL2 FB noise shall be calculated and injected as defined in section A.2.1.2.
a) Testprofile A2_Fix_F_7288k
4 individual tests – 4 shall be passed
© The Broadband Forum. All rights reserved.
89
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-25/TR-100: Fixed rate, ADSL2 FB noise, testprofile A2_Fix_F_7288k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2_Fix_F_7288k
0
750
1250
1500
b) Testprofile A2_Fix_F_5952k
4 individual tests – 4 shall be passed
Table A.2-26/TR-100: Fixed rate, ADSL2 FB noise, testprofile A2_Fix_F_5952k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
A2_Fix_F_5952k
0
750
1250
1850
c) Testprofile A2_Fix_F_2400k
5 individual tests – 5 shall be passed
© The Broadband Forum. All rights reserved.
90
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-27/TR-100: Fixed rate,ADSL2 FB noise, testprofile A2_Fix_F_2400k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
A2_Fix_F_2400k
0
750
1250
2000
2550
d) Testprofile A2_Fix_I_7288k
4 individual tests – 4 shall be passed
Table A.2-28/TR-100: Fixed rate, ADSL2 FB noise, testprofile A2_Fix_I_7288k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
A2_Fix_I_7288k
0
750
1250
1600
e) Testprofile A2_Fix_I_5952k
4 individual tests – 4 shall be passed
© The Broadband Forum. All rights reserved.
91
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-29/TR-100: Fixed rate,ADSL2 FB noise, testprofile A2_Fix_I_5952k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2_Fix_I_5952k
0
750
1250
1950
f) Testprofile A2_Fix_I_2400k
5 individual tests – 5 shall be passed
Table A.2-30/TR-100: Fixed rate,ADSL2 FB noise, testprofile A2_Fix_I_2400
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2_Fix_I_2400
0
750
1250
2000
2650
A.2.6.2 ADSL2plus FB noise impairment
The ADSL2plus FB noise shall be calculated and injected as defined in A.2.1.3.
a) Testprofile A2P_Fix_F_10000k
4 individual tests – 4 shall be passed
© The Broadband Forum. All rights reserved.
92
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-31/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_F_10000k
Downstream
Noise
Margin,
Reported
(dB)
Upstream
Noise
Margin,
Reported
(dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2P_Fix_F_10000k
0
750
1250
1500
b) Testprofile A2P_Fix_F_7288k
4 individual tests – 4 shall be passed
Table A.2-32/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_F_7288k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
A2P_Fix_F_7288k
0
750
1250
1800
c) Testprofile A2P_Fix_F_5952k
4 individual tests – 4 shall be passed
© The Broadband Forum. All rights reserved.
93
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-33/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_F_5952k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2P_Fix_F_5952k
0
750
1250
2000
d) Testprofile A2P_Fix_F_2400k
5 individual tests – 5 shall be passed
Table A.2-34/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_F_2400k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
A2P_Fix_F_2400k
0
750
1250
2000
2550
e) Testprofile A2P_Fix_I_10000k
4 individual tests – 4 shall be passed (applies to Class A implementations only)
© The Broadband Forum. All rights reserved.
94
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-35/TR-100: Fixed rate, ADSL2plus FB noise, testprofile A2P_Fix_I_10000k
Downstream
Noise Margin,
Reported
(dB)
Upstream
Noise Margin,
Reported
(dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2P_Fix_I_10000k
0
750
1250
1600
f) Testprofile A2P_Fix_I_7288k
4 individual tests – 4 shall be passed
Table A.2-36/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_I_7288k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
A2P_Fix_I_7288k
0
750
1250
1900
g) Testprofile A2P_Fix_I_5952k
4 individual tests – 4 shall be passed
© The Broadband Forum. All rights reserved.
95
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-37/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_I_5952k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
A2P_Fix_I_5952k
0
750
1250
2150
h) Testprofile A2P_Fix_I_2400k
5 individual tests – 5 shall be passed
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
A2P_Fix_I_2400k
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
Table A.2-38/TR-100: Fixed rate, ADSL2plus FB noise, testprofile
A2P_Fix_I_2400k
0
750
1250
2000
2650
A.2.7 Loop Tests with Ports Set for ADSL2 Annex L operating mode
The same pass/fail criteria as defined in section A.1-4 apply.
Table A.2-39 lists the number of test points per section corresponding to the 10% and
25% reported margin requirements limits mentioned above.
© The Broadband Forum. All rights reserved.
96
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.2-39/TR-100: Reported Margin Requirements
Number of test cases
Section
number
10% limit
12
12
A.2.7.1
A.2.7.2
1
1
25% limit
(applies to downstream margins only)
2
2
A.2.7.1 White noise impairment
Set noise generator G4 to apply white noise disturber at both ends of the total loop at -140
dBm/Hz.
12 individual tests – 10 tests must be passed
Table A.2-40/TR-100: White noise Impairment ADSL 2Annex L test profile
A2L_RA_I_16000k
A2L_RA_I_16000k
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Downstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
560
504
424
304
240
200
Pass/Fail
Expected
4000
4500
5000
5500
5800
6000
Measured
Loop Length
(m, ETSI loop #1)
Upstream
Sync Rate (kbps)
3360
2296
1480
912
608
400
A.2.7.2 ADSL2 FD noise impairment
FD noise settings as defined in section A.2.1.1 for a loop length of 4500m
12 individual tests – 10 tests must be passed
© The Broadband Forum. All rights reserved.
97
ADSL2/ADSL2plus Performance Test Plan
TR-100
Noise Margin,
Reported (dB)
Pass/Fail
Pass/Fail
Measured
Measured
520
448
416
360
312
280
A2L_RA_I_16000k
Downstream
Sync Rate (kbps)
Expected
Expected
4000
4500
4750
5000
5250
5500
Upstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Loop Length
(m, ETSI loop #1)
Table A.2-41/TR-100: ADSL2 Noise FD 4500 m Impairment, ADSL 2 Annex L test
profile A2L_RA_I_16000k
2096
1248
952
680
448
232
A.2.8 REIN noise impairment
Configure the test setup as specified in section 7.2.2.
The noise models for 19 ADSL2plus disturbers, to be used with the crosstalk coupling
functions defined in section 5.3.2 table 5 of ETSI TS 101 388[8], to generate simulated
crosstalk noise for the corresponding length of ETSI loop #1, are defined in Annex-D
tables D.1-3 and D.1-4.
The crosstalk noise shall be injected at both the ATU-C and ATU-R
In addition to the crosstalk noise at ATU-R, the REIN noise is injected with a repetition
frequency of 100Hz.
Set noise generator G4 to apply white noise disturber at both ends of the total loop at -140
dBm/Hz.
Set noise generator G1 and G2 for 19 FD ADSL2plus noise:
At CO side:
• G1 = X.LT.FD19 for ADSL2plus, with NEXT coupling function.
• G2 = X.NT.FD19 for ADSL2plus, with FEXT coupling function.
At CPE side:
• G1 = X.NT.FD19 for ADSL2plus, with NEXT coupling function.
• G2 = X.LT.FD19 for ADSL2plus, with FEXT coupling function.
The target margin for rate adaptive profile test shall be configured as indicated in table
A.2-42
If an individual test is failed it may be repeated. An individual test may be repeated
several times, but the total number of test attempts required to pass the 8 individual tests
shall not exceed 16.
© The Broadband Forum. All rights reserved.
98
ADSL2/ADSL2plus Performance Test Plan
TR-100
8 individual tests – 7 tests must be passed
Table A.2-42/TR-100: REIN noise impairment, rate adaptive profile
1750
9
3900
2450
9
1900
3150
9
600
Pass / Fail
7644
Measured D/S
Errored Seconds
6
Downstream Noise
Margin, Reported
(dB)
1050
Measured DS sync
rate (kbps)
Target Margin
dB
Expected DS sync
rate
loop length (m,
ETSI_loop #1)
Modem trained and
did not lose sync?
(Y/N)
A2P_RA_I_30000k with max delay=8ms
Table A.2-43/TR-100 : REIN noise impairment, fixed rate profile
1300
A2P_Fix_I_7288k
7288
1500
A2P_Fix_I_5952k
5952
2000
A2P_Fix_I_2400k
2400
2400
A2_Fix_I_1200k
1200
Pass / Fail
Measured D/S
Errored Seconds
Measured
D/S Bit Rate
Downstream
Noise Margin,
Reported (dB)
Modem Trained
and did not loose
sync? (Y/N)
Testprofile, with
delay set to 8 ms.
Loop lenght (m,
ETSI loop #1.)
Testprofile A2(P)_FIX_I_X, with max delay= 8ms
Note: the ADSL2plus disturber noise model is used even when operating in
ADSL2 mode since it can be assumed that the 19 disturbers are on shorter
lines than the long line forced to use ADSL2.
© The Broadband Forum. All rights reserved.
99
ADSL2/ADSL2plus Performance Test Plan
A.3
TR-100
ADSL2plus European tests operating with DS-PSD mask above ADSL
The ADSL2plus over ADSL tests shall be performed with crosstalk noise representing 19
disturbers with the same upstream and downstream PSD as the units under test.
CAL is the Cabinet Assigned Loss. The CAL parameter describes the electrical length of
the cable between the Exchange and the Cabinet.
The tests shall be performed with three different downstream PSD mask shapes,
calculated using the methods described in reference [15] Part B "ANFP PSD Mask
Definition for SDF", for three CAL values: 12, 36, and 52 respectively.
The template PSD’s are given in [16] sheet “Part B”.
The upstream PSD mask shall be independent of the CAL value under test.
The ADSL2plus over ADSL FD noise shall be calculated using the “Definition of the
cross-talk coupling functions” from TS101388[8] section 5.3.2 table 5, for the
corresponding length of ETSI loop #2.
The downstream PSDs and downstream equivalent 19 disturber level X.LT.C12,
X.LT.C36 and X.LT.C52 ADSL2plus over ADSL are shown in Annex-D tables D.2-1,
D.2-2 and D.2-3.
The upstream PSD and the upstream equivalent disturber level X.NT.FD19 ADSL2plus
over ADSL for 19 disturbers are shown in Annex-D table D.1-4.
Set noise generator G4 to apply white noise disturber at both ends of the total loop at -140
dBm/Hz.
Set noise generator G1 and G2 for FD ADSL2plus over ADSL noise:
At CO side:
• G1 = X.LT.Cxx for ADSL2plus over ADSL, for appropriate CAL, with NEXT
coupling function.
• G2 = X.NT.FD19 for ADSL2plus over ADSL, with FEXT coupling function.
At CPE side:
• G1 = X.NT.FD19 for ADSL2plus over ADSL, with NEXT coupling function.
• G2 = X.LT.Cxx for ADSL2plus over ADSL, for appropriate CAL, with FEXT
coupling function.
The same pass/fail criteria as defined in section A.1-4 apply.
Table A.3-1 lists the number of test points per section corresponding to the 10% and 25%
margin requirements limits mentioned above.
© The Broadband Forum. All rights reserved.
100
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.3-1/TR-100: Reported Margin Requirements
Section
number
Number of test cases
25% limit
(applies to downstream margins only)
1
1
4
4
10% limit
10
10
28
28
A.3.1.1
A.3.1.2
A.3.2.1
A.3.2.2
1
1
3
3
A.3.1 Noise ADSL2plus over ADSL FD (rate adaptive)
A.3.1.1 Noise FD ADSL2plus over ADSL Impairment, interleaved profiles
Interleaved profiles shall be tested with testprofile A2P_RA_I_30000k with DelayMax
set to 8ms
10 individual tests – 8 tests must be passed
1000
820
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_RA_I_30000k CAL=12
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #2)
Table A.3-2/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_RA_I_30000k, CAL=12
11460
© The Broadband Forum. All rights reserved.
101
ADSL2/ADSL2plus Performance Test Plan
TR-100
1000
840
7250
2000
860
6100
2400
850
3960
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_RA_I_30000k CAL=36
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #2)
Table A.3-3/TR-100: Noise FD ADSL2plus over ADSL Impairment, testprofile
A2P_RA_I_30000k, CAL=36
Table A.3-4/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_RA_I_30000k, CAL=52
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
880
Pass/Fail
Expected
1000
Measured
Loop Length
(m, ETSI loop #2)
A2P_RA_I_30000k CAL=52
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
11360
A.3.1.2 Noise FD ADSL2plus over ADSL Impairment, fast profiles
10 individual tests – 8 tests must be passed
© The Broadband Forum. All rights reserved.
102
ADSL2/ADSL2plus Performance Test Plan
TR-100
1000
990
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_RA_F_30000k CAL=12
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #2)
Table A.3-5/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_RA_F_30000k, CAL=12
11850
1000
930
6860
2000
960
5720
2400
940
3600
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_RA_F_30000k CAL=36
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #2)
Table A.3-6/TR-100: Noise FD ADSL2plus over ADSL Impairment, testprofile
A2P_RA_F_30000k, CAL=36
103
ADSL2/ADSL2plus Performance Test Plan
TR-100
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
920
Pass/Fail
Expected
1000
Measured
Loop Length
(m, ETSI loop #2)
A2P_RA_F_30000k CAL=52
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Table A.3-7/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_RA_F_30000k, CAL=52
10600
A.3.2 Noise ADSL2plus over ADSL FD (fixed rate)
A.3.2.1 Noise FD ADSL2plus over ADSL impairment, interleaved profiles
The tests specified in tables A.3-11 up to A.3-16 shall be tested with test profile
A2P_Fix_I_X, with X=600k, 2400k and 5952k, and maximum delay set to 8ms.
28 individual tests – 26 tests must be passed
Table A.3-8/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_Fix_I_2400k, CAL=12
2750
352
2400
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
2400
Pass/Fail
352
Measured
1250
Expected
2400
Noise Margin,
Reported (dB)
352
Pass/Fail
Expected
50
Measured
Loop Length
(m, ETSI loop #2)
A2P_Fix_I_2400k CAL=12
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
104
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table A.3-9/TR-100: Noise FD ADSL2plus over ADSL Impairment, testprofile
A2P_Fix_I_X, CAL=36
750
640
5952
2400k
50
352
2400
2400k
750
352
2400
2400k
2250
352
2400
600k
50
128
600
600k
1250
128
600
600k
2500
128
600
Expected
Pass/Fail
Noise Margin,
Reported (dB)
5952k
Pass/Fail
5952
Measured
640
Noise Margin,
Reported (dB)
50
Measured
Loop Length
(m, ETSI Loop #2)
5952k
Expected
TestProfile X
A2P_Fix_I_X CAL=36,with X=600k, 2400k and
5952k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
50
352
2400
1250
352
2400
2750
352
2400
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_Fix_I_2400k CAL=52
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
Table A.3-10/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_Fix_I_2400k, CAL=52
105
ADSL2/ADSL2plus Performance Test Plan
TR-100
A.3.2.2 Noise FD ADSL2plus over ADSL Impairment, fast profiles
28 individual tests – 26 tests must be passed
50
352
2400
1250
352
2400
2750
352
2400
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_Fix_F_2400k CAL=12
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
(m, ETSI loop #2)
Table A.3-11/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_Fix_F_2400k, CAL=12
Table A.3-12/TR-100: Noise FD ADSL2plus over ADSL Impairment, testprofile
A2P_Fix_F_X, CAL=36
640
5952
2400k
50
352
2400
2400k
750
352
2400
2400k
2250
352
2400
600k
50
128
600
600k
1250
128
600
600k
2500
128
600
Expected
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
750
Pass/Fail
5952k
Measured
5952
Expected
640
Noise Margin,
Reported (dB)
50
Pass/Fail
Loop Length
(m, ETSI Loop #2)
5952k
Measured
TestProfile X
A2P_Fix_F_X CAL=36, with X= 600l,2400k and 5952k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
106
ADSL2/ADSL2plus Performance Test Plan
TR-100
50
352
2400
1250
352
2400
2750
352
2400
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
A2P_Fix_F_2400k CAL=52
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Expected
Loop Length
Table A.3-13/TR-100: Noise FD ADSL2plus over ADSL Impairment testprofile
A2P_Fix_F_2400k, CAL=52
107
ADSL2/ADSL2plus Performance Test Plan
TR-100
Annex B Physical Layer Test Cases for systems using
G.992.3/5 Annex B.
This Annex contains a set of tests and requirements specific for ADSL systems according
to the recommendations:
• G.992.3[1] for ADSL2 over ISDN with Frequency Division Duplexing (FDD).
• G.992.5[2] for ADSL2plus over ISDN with Frequency Division Duplexing
(FDD).
Testplan details, especially performance tests, for ADSL over ISDN Systems using echo
cancellation are actually not considered and are for further study.
The performance on short loops is based on the ADSL-like (G.992.1) power cutback
mechanism.
Two performance classes based on the support of optional extended framing parameters
for the use of high delay profile with INP = 2 are defined.
Class A: The DSLAM – modem combination does support the optional extended
framing parameters acc. G.992.3[1] and G.992.5[2]
Class B: The DSLAM – modem combination does not support the optional
extended framing parameters acc. G.992.3[1] and G.992.5[2]
The performance tests have to be performed once and after filling in the results table the
DSLAM – modem combination is allocated to the one or the other performance class.
B.1
Annex B specific test setup information
In addition to the general part of this document the following configurations have to be
considered.
B.1.1 Splitter model
All tests shall be done with splitters inserted in the line on both the ATU-C and ATU-R
side. The vendor of the EUT (DSLAM, MODEM) shall deliver together with the EUT a
splitter that complies with the common recommendations ETSI TS101952-1[9] , SubPart 3: specification of ADSL/ISDN splitters and/or ETSI TS101952-1 [9], Sub-Part 4:
ADSL/’ISDN or POTS’ universal splitter. The splitter manufacturer and design (external,
integrated or splitter line card) is left to the EUT vendor.
The functional model of the used splitters shall conform to the following figure:
© The Broadband Forum. All rights reserved.
108
ADSL2/ADSL2plus Performance Test Plan
TR-100
Figure B.1: structure of the ADSL/ISDN and ADSL/’ISDN or POTS’
universal splitter
This structure of the splitter is a proposal for an external splitter. If the splitter
functionality is integrated into the linecard, then the blocking capacitors are not
mandatory.
B.1.2 Test loops / Loop characteristics
The tests of the physical layer are based on TS101388[8] for both ADSL2 and
ADSL2plus. For the test setup figure 10 applies. The CPE and the DSLAM are connected
via a loop simulator. The loop simulators shall simulate the loop characteristics specified
in the above standards. The loop simulators shall be calibrated relative to the nominal
attenuation as defined in the section 3.1.1.
B.1.3 Noise and noise injection
Noise shall be injected as specified in ETSI TS101388[8], with simultaneous noise
injection at both ends of the loop. Possible deviations from this are stated in the test
cases.
The noise injection shall be calibrated as defined in ETSI TS101388[8].
The simultaneous used generators shall be G1 (NEXT noise), G2 (FEXT noise), G4
(white noise) and G8 (Same Pair ISDN/line sharing noise).
For white noise set generator G4 at both ends of the total loop at –140 dBm/Hz.
For the line sharing noise generator G8 according to TS101388[8] is used.
If G8 is implemented in the above mentioned noise generator, through the noise injection
circuit, the ISDN port of the splitter has to be terminated with the appropriate ISDN
impedance.
If G8 is not implemented in the above mentioned noise generator, G8 has to be realized
according to TS101388[8], Annex F. In this case the ISDN port of the splitter has to be
used for injecting the line sharing noise.
B.1.3.1 Noise Generator settings G1,d G2 and G4 for ADSL2, Annex B
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
The noise model for G1 and G2 shall be FB.:
At CO side:
© The Broadband Forum. All rights reserved.
109
ADSL2/ADSL2plus Performance Test Plan
TR-100
•
G1 = X.LT.FB as defined in TS101388[8], section 5.3.4.1.4, table 14, with NEXT
coupling function
• G2 = X.NT.FB as defined in TS101388[8], section 5.3.4.1.4, table 15, with FEXT
coupling function
At CPE side:
• G1 = X.NT.FB as defined in TS101388[8], section 5.3.4.1.4, table 15, with NEXT
coupling function
• G2 = X.LT.FB as defined in TS101388[8], section 5.3.4.1.4, table 14, with FEXT
coupling function
All measurements with noise model FB shall be done with the noise model FB calculated
for 2875m loop #1 (ETSI performance requirement for 768 kbps downstream and 128
kbps upstream, see TS101388[8], tables 38 and 39 for FDD ADSL over ISDN down- and
upstream). To reduce test time only one noise model is used for all measurement points.
B.1.3.2 Noise Generator settings G1, G2 and G4 for ADSL2, Annex B noise FA or
FD impairment,
All measurements with noise model FA and FD shall be done with the noise model
calculated for the corresponding length of the ETSI loop selected as indicated in the
applicable testsection.
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
Set noise generator G1 and G2 for ETSI FA/FD noise:
At CO side:
• G1 = X.LT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 15,
with NEXT coupling function.
• G2 = X.NT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 14,
with FEXT coupling function.
At CPE side:
• G1 = X.NT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 14,
with NEXT coupling function.
• G2 = X.LT.FA/FD as defined in TS 101 388 [8], section. 5.3.4.1.3, table 15,
with FEXT coupling function.
B.1.3.3 Noise Generator settings G1,G2 and G4 for ADSL2plus,Annex B noise FB
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
The noise model for G1 and G2 shall be ADSL2plus Annex-B FB:
At CO side:
• G1 = X.LT.FB for ADSL2plus as defined in Annex D.3 table D.3-2, with NEXT
coupling function
• G2 = X.NT.FB for ADSL2plus as defined in Annex D.3 table D.3-1, with FEXT
coupling function
At CPE side:
© The Broadband Forum. All rights reserved.
110
ADSL2/ADSL2plus Performance Test Plan
•
•
TR-100
G1 = X.NT.FB for ADSL2plus as defined in Annex D.3 table D.3-1, with NEXT
coupling function
G2 = X.LT.FB for ADSL2plus as defined in Annex D.3 table D.3-2, with FEXT
coupling function
All measurements with noise model FB shall be done with the noise model FB calculated
for the corresponding length of ETSI loop #1.
B.1.3.4 Noise Generator settings G1, G2 and G4 for ADSL2plus, Annex B noise FA
or FD impairment
Set noise generator G4 to apply a white noise disturber at both ends of the loop at -140
dBm/Hz.
All measurements with noise model FA/FD shall be done with the noise model FA/FD
calculated for the corresponding length of the ETSI loop selected as indicated in the
applicable testsection.
The noise model for G1 and G2 shall be ADSL2plus Annex-B FA or FD:
At CO side:
• G1 = X.LT.FA/FD for ADSL2plus as defined in Annex D.3 table
NEXT coupling function
• G2 = X.NT.FA/FD for ADSL2plus as defined in Annex D.3 table
FEXT coupling function
At CPE side:
• G1 = X.NT.FA/FD for ADSL2plus as defined in Annex D.3 table
NEXT coupling function
• G2 = X.LT.FA/FD for ADSL2plus as defined in Annex D.3 table
FEXT coupling function
D.3-2, with
D.3-1, with
D.3-1, with
D.3-2, with
B.1.4 Test procedure
Tests shall be performed according to the general procedure described in section 7.
In addition to achieving the required rate, both downstream and upstream noise margin
values are to be considered in determining the result of an individual section.
It is acknowledged that achieving a desired noise margin is primarily the responsibility of
the receiver. That is, the DSLAM is primarily responsible for achieving desired upstream
noise margins, while the CPE (modem) is primarily responsible for achieving desired
downstream noise margins.
Table B.1-1 outlines the Noise margin Pass/Fail criteria.
© The Broadband Forum. All rights reserved.
111
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.1-1/TR-100: Noise margin chart
Reported Noise
Margin (dB)
<5
>= 5 and < 6
>= 6
Requirement
On no test point
On at most 10% of the test
points
On at least 90% of the test
points
All values shall be read out at the DSLAM.
Violation of any of the requirements in the Noise Margin Chart shall constitute a test
section failure.
Overall pass/fail criteria for each adaptive rate test are then as follows:
• If any reported noise margin is less than 5 dB, then the ATU-C/ATU-R pair
fails the noise margin requirements of that section.
• If more than 10% of the reported noise margins are less than 6dB in a section,
then the ATU-C/ATU-R pair fails the noise margin requirements of that
section.
• If more than 10% of the data rates are less than the data rate requirements in a
section, then the ATU-C/ATU-R pair fails the data rate requirements of that
section.
• If the ATU-C/ATU-R pair passes both the data rate and noise margin
requirements, it passes the section; otherwise, it fails the section.
Overall pass/fail criteria for each fixed rate test section are then as follows:
• If any reported noise margin is less than 6 dB, then the ATU-C/ATU-R pair
fails the noise margin requirements of that section.
• If any of the data rates are less than the data rate requirements in a section,
then the ATU-C/ATU-R pair fails the data rate requirements of that section.
• If the ATU-C/ATU-R pair passes both the data rate and noise margin
requirements, it passes the section; otherwise, it fails the section.
Table B.1-2 lists the number of test points per section corresponding to the overall
pass/fail criteria (10% limit).
© The Broadband Forum. All rights reserved.
112
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.1-2/TR-100: 10% limit for Annex B.2 test sections
Section
Number
B.2.4.1
B.2.4.2
B.2.5.1
B.2.5.2
B.2.5.3 a)
B.2.5.3 b)
Number of test
points in section
36
28
34
30
0 ≤ x ≤ 10
0 ≤ x ≤ 10
10% limit
4
3
n/a
n/a
1
1
Table B.1-3 lists the number of test points per section corresponding to the overall
pass/fail criteria (10% limit) mentioned in section B.1.4.
Table B.1-3/TR-100: 10% limit for Annex B.3 test sections
Section
Number
B.3.4.1
B.3.4.2
B.3.5.1
B.3.5.2
B.3.5.3 a)
B.3.5.3 b)
B.3.7
B.3.8
B.2
Number of test
points in section
36
28
44
38
0 ≤ x ≤ 10
0 ≤ x ≤ 10
6
5
10% limit
4
3
n/a
n/a
1
1
1
1
Performance European Test set for ADSL2, Annex B
The specific testprofiles as defined in table 7.3 shall be used. The test case dependent
parameters are defined as follows:
Table B.2-1/TR-100: testcase dependant parameters for ADSL2, Annex-B
Parameter
CARMASKds
Setting
1-59
CARMASKus
1 – 32 / 60-63
Description
mask on: 1 –59
mask off: 60-255 (used downstream
subcarriers)
Ö First-last subcarrier for ds: 60-255
Ö First-last subcarrier for us: 33-59
© The Broadband Forum. All rights reserved.
113
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.2.1 CPE Margin verification tests
For a detailed description of the margin verification test methodology/procedure, see
section 7.5
The following tables B.2-2 up to B.2-8 show the loop and noise conditions for margin
verification tests.
B.2.1.1 Noise ADSL2 FB impairments at 750m.
Settings as defined in section B.1.3.1, using testprofile B2_RA_F_16000.
Table B.2-2/TR-100: Noise ADSL2 FB Impairment, testprofile B2_RA_F_16000k at
750 m
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated DS
net data rate:
>= 7544 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test end)
750
Estimated BER,
based on equations
in Table 7-12, fast
path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
5
B.2.1.2 Noise ADSL2 FB impairment at 2750m.
Settings as defined in section B.1.3.1, using testprofile B2_RA_I_16000k.
© The Broadband Forum. All rights reserved.
114
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-3/TR-100: Noise ADSL2 FB impairment, testprofile B2_RA_I_16000k at
2750 m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated DS
net data rate:
>= 1856 kbps
Achieved DS
net data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
2750
Estimated BER,
based on equations
in Table 7-12,
interleaved path
40
B.2.1.3 Random selected loop 1 (Optional test)
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2 as defined in section B.1.3.2, using testprofile B2_RA_I_16000k.
Table B.2-4/TR-100: First Selected Test Loop, testprofile B2_RA_I_16000k
Test time
ETSI
(minutes)
Loop
Selected Calculate
based on
connect
rate.
Achieved DS
net data rate
(test start)
Achieved DS
net data rate
(test end)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Estimated BER,
based on equations
in Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Loop
Selected
Note: The table B.2-4/TR-100 is an example table. The specific test loop
selected and noise shape used should be indicated in the table. The connect
rate and also test time calculated shall be indicated in the table B.2-4.
B.2.1.4 Random selected loop 2 (Optional test)
© The Broadband Forum. All rights reserved.
115
ADSL2/ADSL2plus Performance Test Plan
TR-100
Select one test loop, different from the loop selected in section B.2.1.3, from: ETSI Loop
#3-7, with either Noise Impairment FA or FD for ADSL2 as defined in section B.1.3.2,
using testprofile B2_RA_I_16000k.
Table B.2-5/TR-100: Second Selected Test Loop, testprofile B2_RA_I_16000k
ETSI
Loop
Selected
Test time
(minutes)
Calculate
based on
connect
rate.
Achieved DS net
data rate
(test start)
Measured DS CRC count, after
injected noise level has been raised
by Min(initial_reported_margin 1, target_margin) dB
Achieved DS net
data rate
(test end)
Estimated BER, based
on equations in Table
Pass
/
Fail
7-12, interleaved
path
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Loop
Selected
Note: The table B.2-5 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and test time
calculated shall also be indicated in the table B.2-5.
B.2.2 CO Margin Verification (Optional)
B.2.2.1 Noise ADSL2 FB impairments at 750m.
Noise FB settings as defined in section B.1.3.1, using testprofile B2_RA_F_16000k.
Table B.2-6/TR-100: Upstream Margin Test with Noise ADSL2 FB Impairment,
testprofile B2_RA_F_16000k at 750m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated US
net data rate:
>= 640 kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test end)
750
Estimated BER,
based on equations
in Table 7-12, fast
path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
40
© The Broadband Forum. All rights reserved.
116
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.2.2.2 Noise ADSL2 FB impairments at 2750m
Noise FB settings as defined in section B.1.3.1, using testprofile B2_RA_I_16000k.
Table B.2-7/TR-100: Upstream Margin Test with Noise ADSL2 FB Impairment,
testprofile B2_RA_I_16000k, at 2750m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated US
net data rate:
>= 224kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test end)
2750
Estimated BER,
based on equations
in Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER <
1.75e-7
then PASS
else FAIL
150
Note: Due to the low, achievable DS data rates under this loop and noise
condition, the observation of 10 error events would require approximately 300
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of
observed error events is reduced to 5. To remain consistent with previous
confidence levels of estimated BER, the range of allowed estimated BER is
increased from 1.5-e7 to 1.75e-7.
B.2.2.3 Random selected loop 1
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2 as defined in section B.1.3.2, using testprofile B2_RA_I_16000k.
© The Broadband Forum. All rights reserved.
117
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-8/TR-100: Downstream Margin Tests for First Selected ETSI Loop #3 – 7,
testprofile B2_RA_I_16000k.
ETSI
Test
Loop
time
Selected (minutes)
Calculate
based on
connect
rate.
Achieved
DS net
data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin -1,
target_margin) dB
Achieved
DS net data
rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
Loop
Selected
Note: the table B.2-8 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and test time
calculated shall also be indicated in the table B.2-8.
B.2.2.4 Random selected loop 2
Select one test loop, different from the loop selected in section B.2.2.3, from: ETSI Loop
#3-7 with either Noise Impairment FA or FD for ADSL2 as defined in section B.1.3.2,
using testprofile B2_RA_I_16000k.
Table B.2-9/TR-100: Downstream Margin Tests for Second Selected ETSI Loop #3
– 7, testprofile B2_RA_I_16000k.
Test
ETSI
time
Loop
Selected (minutes)
Calculate
based on
connect
rate.
Achieved
DS net
data rate
(test start)
Measured DS CRC count,
after injected noise level
has been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
DS net data
rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER <
1.5e-7
then PASS
else FAIL
Loop
Selected
Note: the table B.2-9 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and test time
calculated shall also be indicated in the table B.2-9.
© The Broadband Forum. All rights reserved.
118
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.2.3 Verification of Downstream bi/gi Values
Table B.2-9 describes the testprocedure for the verification of downstream bi/gi values in
ADSL2 operating mode.
Table B.2-10/TR-100: Verification of downstream bi/gi values
Test
Configuration
The ATU-C and ATU-R are connected in turn through the
following loops:
Method of
Procedure
[1] ETSI loop #1, 750 m with -140 dBm/Hz AWGN, testprofile
B2_RA_F_16000k
[2] ETSI loop #1, 2750 m with -140 dBm/Hz AWGN, testprofile
B2_RA_I_16000k
[3] ETSI loop #1, 750 m with noise FB as defined in section B.1.3.1,
testprofile B2_RA_I_16000k
[4] ETSI loop #1, 2750 m with noise FB as defined in section
B.1.3.1, testprofile B2_RA_F_16000k
[1] Select test loop 1 from the set above.
[2] Train the modem in the test profile defined for the selected loop.
[3] From the DSLAM read the bi and gi values requested during
initialisation
[4] Repeat steps 1 to 3 for other test loops in the set.
The RMSGIds shall be calculated in linear scale with gi values as
contained in the BiGi table and then rounded to nearest 100th of a dB.
The bi and gi values rounded to the nearest 100th of a dB in the bitsand-gains table shall comply with the following requirements
(Sections 8.5.2 and 8.6.4 of G.992.3[1])
Expected Results
[1] All bi values shall be in the [0 to BIMAXds] (bits) range, where 8
≤ BIMAXds ≤ 15.
[2] If bi > 0, then gi shall be in the [–14.54 to +2.50 + EXTGIds]
(dB) range where 0 ≤ EXTGIds ≤ MAXNOMPSDds –
NOMPSDds.
[3] RMSGIds ≤ EXTGIds + 1/16th of a dB
[4] If bi > 0, then gi shall be in the [RMSGIds – 2.60 to RMSGIds +
2.60] (dB) range.
[5] If bi = 0, then gi shall be equal to 0 (linear) or in the [–14.54 to
RMSGIds + 0.10] (dB) range.
BIMAXds, EXTGIds, MAXNOMPSDds, NOMPSDds, RMSGIds
are defined in G.992.3[1] Sections 8.5.1 and 8.5.2.
B.2.4 Loop Tests with Ports Set for Adaptive Rate
B.2.4.1 White Noise Impairment
Set noise generator G4 to apply white noise disturber at both ends of the total loop at –
140 dBm/Hz.
© The Broadband Forum. All rights reserved.
119
ADSL2/ADSL2plus Performance Test Plan
TR-100
18 individual tests – 16 tests must be passed
Table B.2-11/TR-100: White noise impairment, testprofile B2_RA_F_16000k
0
1096
10000
500
1096
10000
1000
1096
10000
1500
1096
10000
2000
1048
8656
2500
960
6560
3000
744
4064
3500
544
2200
4000
344
720
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Downstream
Sync Rate (kbps)
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length [m,
loop#1]
B2_RA_F_16000
18 individual tests – 16 tests must be passed
Noise Margin,
Reported (dB)
Pass/Fail
Pass/Fail
Measured
Measured
888
888
888
888
888
888
776
584
384
B2_RA_I_16000k
Downstream
Sync Rate (kbps)
Expected
Expected
0
500
1000
1500
2000
2500
3000
3500
4000
Upstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Loop Length [m,
loop#1]
Table B.2-12/TR-100: White noise impairment, testprofile B2_RA_I_16000k
7608
7608
7608
7608
7608
6872
4584
2536
1000
© The Broadband Forum. All rights reserved.
120
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.2.4.2 Noise FB impairment
Set noise generators G1, G2, G4 and G8 for model FB impairment as defined in section
B.1.3.
Note: Short loop power cutback in ADSL2 can be determined on request by
either side (DSLAM and CPE), depending on the receiver front-end
capabilities. I.e. a short loop power cutback requested by CPE determines
downstream aggregate transmit power, which mainly impacts downstream
performance. Downstream performance targets defined in Table B.2-12 and
B.2-13 take into account a maximum power cutback of 12 dB.
14 individual tests – 12 must be passed
1250
800
7251
1750
600
5963
2250
392
3786
2750
176
1532
3000
64
400
Pass/Fail
7551
Measured
960
Expected
750
Noise Margin,
Reported (dB)
6470
Pass/Fail
1024
Measured
250
Expected
Loop Length [m,
loop#1]
B2_RA_F_16000k
Downstream
Sync Rate (kbps)
Upstream
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Table B.2-13/TR-100: Noise FB impairment, testprofile B2_RA_F_16000k
14 individual tests – 12 must be passed
© The Broadband Forum. All rights reserved.
121
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-14/TR-100: Noise FB impairment, testprofile B2_RA_I_16000k
250
888
6434
750
888
7508
1250
824
7334
1750
640
6165
2250
432
4121
2750
224
1861
3000
96
640
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Downstream
Sync Rate (kbps)
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Upstream
Sync Rate (kbps)
Expected
Loop Length [m,
loop#1]
B2_RA_I_16000k
B.2.5 Loop Tests with Ports Set For Fixed Rate
B.2.5.1 White Noise Impairment
Set noise generator G4 to apply white noise disturber at both ends of the total loop
at –140 dBm/Hz.
a) Rate Profile: B2_Fix_F_864k and B2_Fix_I_864
12 individual tests – 12 shall be passed
© The Broadband Forum. All rights reserved.
122
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-15/TR-100: White noise fixed rate profile 864k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
B2_Fix_I_864k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
B2_Fix_F_864k
0
500
1000
2000
3000
Max. loop length requirement for
fast mode, test not to perform
3900
4000
Max. loop length requirement for
interleaved mode, test not to perform
b) Rate Profile: B2_Fix_F_3456k and B2_Fix_I_3456k
12 individual tests – 12 shall be passed
Table B.2-16/TR-100: White noise fixed rate profile 3456k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin, Reported
(dB)
Modem Trained
(Y/N)?
B2_Fix_I_3456k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin, Reported
(dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2_Fix_F_3456k
0
500
1000
2000
2750
Max. loop length requirement for
fast mode, test not to perform
3150
3200
Max. loop length requirement for
interleaved mode, test not to perform
c) Rate Profile: B2_Fix_F_7288k and B2_Fix_I_7288k
© The Broadband Forum. All rights reserved.
123
ADSL2/ADSL2plus Performance Test Plan
TR-100
10 individual tests – 10 shall be passed
Table B.2-17/TR-100: White noise fixed rate profile 7288k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
B2_Fix_I_7288k
Modem
Trained (Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
B2_Fix_F_7288k
0
500
1000
2000
Max. loop length requirement for
low delay, test not to perform
2300
2350
Max. loop length requirement for
high delay, test not to perform
B.2.5.2 European noise FB Impairment
Set noise generators G1, G2, G4 and G8 for model FB impairment, same pair ISDN and
white noise, as defined in section B.1.3.
a) Rate Profile: B2_Fix_F_864k and B2_Fix_I_864k
12 individual tests – 12 shall be passed
© The Broadband Forum. All rights reserved.
124
ADSL2/ADSL2plus Performance Test Plan
TR-100
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained
(Y/N)?
B2_Fix_I_864k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
B2_Fix_F_864k
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
Table B.2-18/TR-100: Noise FB impairment fixed rate profile 864k
0
500
1000
2000
2500
Max. loop length requirement for
fast mode, test not to perform
2750
2850
Max. loop length requirement for
interleaved mode, test not to perform
b) Rate Profile: B2_Fix_F_3456k and B2_Fix_I_3456k
10 individual tests – 10 shall be passed
Table B.2-19/TR-100: Noise FB impairment fixed rate profile 3456k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
B2_Fix_I_3456k
Modem Trained
(Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2_Fix_F_3456k
0
500
1000
1750
Max. loop length requirement for
fast mode, test not to perform
2100
2150
Max. loop length requirement for
interleaved mode, test not to perform
c) Rate Profile: B2_Fix_F_7288k and B2_Fix_I_7288k
8 individual tests – 8 shall be passed
© The Broadband Forum. All rights reserved.
125
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-20/TR-100: Noise FB impairment fixed rate profile 7288k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
B2_Fix_I_7288k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
B2_Fix_F_7288k
500
750
1000
Max. loop length requirement for
fast mode, test not to perform
1150
Max. loop length requirement for
interleaved mode, test not to perform
Note: because of the lower performance on short loops, the test starts at 500m.
1200
B.2.5.3 Target Noise margin consideration
To ensure the consideration of the desired target noise margin for fixed rate operation
during/after synchronization the noise margin for the maximum loop length for the below
described scenario has to be recorded.
The defined requirement is a CPE connecting with the target noise margin. For this test
the maximum loop length achieved is not a pass/fail consideration.
a) Target noise margin consideration upstream
© The Broadband Forum. All rights reserved.
126
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-21/TR-100: Target noise margin consideration upstream
[1] Rate parameters acc. B2_Fix_F_3456k.
[2] Set noise generators for model FB impairment as defined in
section B1.3.
Method of Procedure [1] Start test at maximum length from test B.2.5.2 b) at which the
modem was able to train successfully
[2] Allow to train for 60 seconds to achieve showtime.
[3] Record the upstream and downstream reported noise margin in
the table below. (Read the margin one minute after show time
is reached.)
[4] If margin upstream > target noise margin + 1 dB, increase the
loop length by:
next loop length step =
floor[(reported noise margin – target noise margin)]/0.6*50m
[5] Repeat steps 2 – 3, if the modem was able to train increase the
loop length 50 meters and follow steps 5, until the modem fails
to connect.
[6] After the first length for which the modem fails to connect
attempt the next incremented loop length. (50 meters greater
then the failing loop length of step 5.) If the CPE connects
record the reported upstream and downstream margin.
Expected Result to If the modem reaches showtime for a certain loop length, then the
related reported noise margin must meet the pass/fail criteria for the
Pass
rate adaptive tests (10% limit) as stated in section B.1.4.
Test Configuration
Table B.2-22/TR-100: Target noise margin consideration upstream - Results table
Pass/Fail
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin
Reported
(dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
B2_Fix_F_3456k
note 1
..
2250
2300
2350
…
…
Note 1: max. length from test B.2.5.2 b), at which the modem was able to train successfully
b) Target noise margin consideration downstream
© The Broadband Forum. All rights reserved.
127
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.2-23/TR-100: Target noise margin consideration downstream
[1] Rate parameters acc. B2_Fix_F_3456k.
[2] Set noise generators G4 to apply white noise impairment as
defined in section B1.3.
Method
of [1] Start test at maximum length from test B.2.5.1 b) at which
the modem was able to train successfully
Procedure
[2] Allow to train for 60 seconds to achieve showtime.
[3] Record the upstream and downstream reported noise
margin in the table below. (Read the margin one minute
after show time is reached.)
[4] If margin downstream > target noise margin + 1 dB,
increase the loop length by: next loop length step =
floor[(reported
noise
margin
–
target
noise
margin)/1.0]*50m
[5] Repeat steps 2 – 3, if the modem was able to train increase
the loop length 50 meters and follow steps 5, until the
modem fails to connect.
[6] After the first length for which the modem fails to connect
attempt the next incremented loop length. (50 meters
greater then the failing loop length of step 5.) If the CPE
connects record the reported upstream and downstream
margin.
Expected Result to If the modem reaches showtime for a certain loop length, then
the related reported noise margin must meet the pass/fail
Pass
criteria as stated in section B.1.4.
Test Configuration
Table B.2-24/TR-100: Target noise margin consideration downstream - Results
table
Pass/Fail
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin
Reported
(dB)
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
B2_Fix_F_3456k
note 2
..
3275
3325
3375
…
…
Note 2: max. length from test B.2.5.1 b), at which the modem was able to train
successfully
© The Broadband Forum. All rights reserved.
128
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.2.6 Performance in L2 mode
With this test it shall be proven that after a transition to L2 and back to L0, the
performance (noise margin) is at least the same as in the initial L0 stage. For this test no
change in loop/noise conditions are to be done during the test duration. This means the
test should show that the power management transitions have no impact on the
performance.
Table B.2-25/TR-100: ADSL2 performance in L2 mode
In addition to the configuration of section B.1 the following has to
be defined:
[1] Test configuration: Figure 4 to Figure 6, depending on
customer interface of modem.
[2] Configure the line with testprofile B2_RA_I_16000k (note 1).
[3] PMMode: Low power state allowed (bit 0 = 0, bit 1 = 1).
[4] PM-Parameters:
i. Minimum (Net) Data rate DS in Low Power
state(L2Min): 128 kbps
ii. L0-TIME:
60 sec.
iii. L2-TIME:
15 sec.
iv. L2-ATPR: 3 dB
v. L2-ATPRT: 18 dB
[5] Set loop simulator to: 2000 m with ETSI loop #1.
[6] Set noise generators for model ADSL2 FB impairment as
defined in section B1.3.1.
Method
of [1] Traffic generator/analyzer switched off or 120kbps DS traffic
[2] Synchronize the line with the required parameters and note the
Procedure
actual downstream sync rate and noise margin (first L0 noise
margin) value.
[3] Wait for automatic transition to L2 with power trims. If
automatic transition does not occur within 120 seconds, force
L2 entry via the G997.1[4] PMFS MIB element.
[4] After 5 minutes: read the actual downstream noise margin
value (L2 noise margin) and DS Tx power (ACTATPds)
[5] Switch on traffic generator with 90% of DS sync rate, to exit
from L2 state to L0 state.
[6] Wait 60 seconds and read the actual DS noise margin (second
L0 noise margin) value.
Expected Result [1] First L0 noise margin >= 5 dB.
[2] L2 noise margin >= 5 dB.
to Pass
[3] Second L0 noise margin >= first L0 noise margin –1 dB.
[4] No CRC error during the entire test.
Test
Configuration
Note 1: The B2_RA_I_16000k minimum net data rate in the L2 mode
test shall be equal to L2_min.
© The Broadband Forum. All rights reserved.
129
ADSL2/ADSL2plus Performance Test Plan
B.3
TR-100
Performance Test set for ADSL2plus, Annex B
The specific test profiles as defined in table 7.3 shall be used . The test case dependent
parameters are defined as follows:
Table B.3-1/TR-100: testcase dependant parameters for ADSL2plus, Annex-B
Parameter
CARMASKds
Setting
1-59
CARMASKus
1 – 32 / 60-63
Description
mask on: 1 –59
mask off: 60-511 (used downstream
subcarriers)
Ö First-last subcarrier for ds: 60-511
Ö First-last subcarrier for us: 33-59
B.3.1 CPE Margin verification tests
Section 7.5 gives a description and testprocedure for the margin verification tests in
tables B.3-2 up to B.3-10
B.3.1.1 Noise FB ADSL2plus impairments at 250m.
Settings as defined in section B.1.3.3, using testprofile B2P_RA_F_30000k
Table B.3-2/TR-100: Noise FB ADSL2plus impairment, testprofile
B2P_RA_F_30000k, at 250m
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 13000
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Achieved DS
net data rate
(test end)
250
Estimated BER,
based on
equations in
Table 7-12, fast
path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
5
B.3.1.2 Noise FB ADSL2plus impairments at 1750m.
Settings as defined in section B.1.3.3, using testprofile B2P_RA_I_30000k
© The Broadband Forum. All rights reserved.
130
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-3/TR-100: Noise FB ADSL2plus impairment, testprofile
B2P_RA_I_30000k, at 1750m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 7640
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Pass
/
Fail
If:
Estimated_BER
< 1. 5e-7
then PASS
else FAIL
Achieved DS
net data rate
(test end)
1750
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
10
B.3.1.3 Noise FB ADSL2plus impairments at 750m.
Settings as defined in section B.1.3.3, using testprofile B2P_RA_F_30000k as defined in
section B.3.8
Table B.3-4/TR-100: Noise FB ADSL2plus impairment, testprofile
B2P_RA_F_30000k, at 750m.
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
DS net data
rate:
>= 5536
kbps
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved DS
net data rate
(test start)
Achieved DS
net data rate
(test end)
750
Estimated BER,
based on
equations in
Table 7-12, fast
path
Pass
/
Fail
If:
Estimated_BER
< 1. 5e-7
then PASS
else FAIL
5
B.3.1.4 Random selected loop 1 (Optional test)
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2plus as defined in section B.1.3.4, using testprofile B2P_RA_I_30000k
© The Broadband Forum. All rights reserved.
131
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-5/TR-100: Downstream Margin Tests for First Selected ETSI Loop #3 – 7,
testprofile B2P_RA_I_30000k
ETSI
Test time
Loop
(minutes)
Selected Calculate
based on
connect
rate.
Achieved
DS net
data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
DS net
data rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
Loop
Selected
Note: the table B.3-5 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and also test
time calculated shall be indicated in the table B.3-5.
B.3.1.5 Random selected loop 2 (Optional test)
Select one other test loop, different from the one selected in section B.3.1.4, from: ETSI
Loop #3-7 with either Noise Impairment FA or FD for ADSL2plus as defined in section
B.1.3.4, using testprofile B2P_RA_I_30000k (Optional test)
Table B.3-6/TR-100: Downstream Margin Tests for Second Selected ETSI Loop #3
– 7, testprofile B2P_RA_I_30000k
Test time
ETSI
(minutes)
Loop
Selected Calculate
based on
connect
rate.
Achieved
DS net
data rate
(test start)
Measured DS CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
DS net
data rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
Loop
Selected
Note: the table B.3-6 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and also test
time calculated shall be indicated in the table B.3-6.
© The Broadband Forum. All rights reserved.
132
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.3.2 CO Margin Verification (Optional)
B.3.2.1 Noise FB ADSL2plus impairments at 250m.
Settings as defined in section B.1.3.3, using testprofile B2P_RA_F_30000k
Table B.3-7/TR-100: Upstream Margin Test with Noise FB ADSL2plus
Impairment, testprofile B2P_RA_F_30000k, at 250 meters
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 1152
kbps
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved US
net data rate
(test start)
Achieved US
net data rate
(test end)
250
Estimated BER,
based on
equations in
Table 7-12, fast
path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
25
B.3.2.2 Noise FB ADSL2plus impairments at 1750m.
Settings as defined in section B.1.3.3, using testprofile B2P_RA_I_30000k
© The Broadband Forum. All rights reserved.
133
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-8/TR-100: Upstream Margin Tests with Noise FB ADSL2plus
Impairment at 1750 meters, testprofile B2P_RA_I_30000k
Loop Test time
length (minutes)
(m)
0.4mm
Anticipated
US net data
rate:
>= 640 kbps
Achieved US
net data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Pass
/
Fail
If:
Estimated_BER
< 1.75e-7
then PASS
else FAIL
Achieved US
net data rate
(test end)
1750
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
55
Note: Due to the low, achievable DS data rates under this loop and noise
condition, the observation of 10 error events would require approximately 105
minutes (at BER ~ 1e-7). To accelerate testing, the desired number of
observed error events is reduced to 5. To remain consistent with previous
confidence levels of estimated BER, the range of allowed estimated BER is
increased from 1.5-e7 to 1.75e-7.
B.3.2.3 Random selected loop 1.
Select one test loop from: ETSI Loop #3-7 with either Noise Impairment FA or FD for
ADSL2plus as defined in section B.1.3.4, using testprofile B2P_RA_I_30000k
Table B.3-9/TR-100: Upstream Margin Tests for First Selected ETSI Loop #3 – 7,
testprofile B2P_RA_I_30000k
Test time
ETSI
(minutes)
Loop
Selected Calculate
based on
connect
rate.
Achieved
US net
data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
US net
data rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
Loop
Selected
© The Broadband Forum. All rights reserved.
134
ADSL2/ADSL2plus Performance Test Plan
TR-100
Note: the table B.3-9 is an example table. The specific test loop selected and
noise shape used shall be indicated in the table. The connect rate and test time
calculated shall also be indicated in the table B.3-9.
B.3.2.4 Random selected loop 2.
Select one other test loop, different from the one selected in section B.3.2.3, from: ETSI
Loop #3-7 with either Noise Impairment FA or FD for ADSL2plus as defined in section
B.1.3.4, using testprofile B2P_RA_I_30000k
Table B.3-10/TR-100: Upstream Margin Tests for Second Selected ETSI Loop #3 –
7, testprofile B2P_RA_I_30000k
ETSI
Test time
Loop
(minutes)
Selected Calculate
based on
connect
rate.
Achieved
US net
data rate
(test start)
Measured US CRC count,
after injected noise level has
been raised by
Min(initial_reported_margin
-1, target_margin) dB
Achieved
US net
data rate
(test end)
Estimated BER,
based on
equations in
Table 7-12,
interleaved path
Pass
/
Fail
If:
Estimated_BER
< 1.5e-7
then PASS
else FAIL
Loop
Selected
Note: the table for section B.3-10 is an example table. The specific test loop
selected and noise shape used shall be indicated in the table. The connect rate
and test time calculated shall also be indicated in the table B.3-10.
B.3.3 Verification of downstream bi/gi values
Table B.3-11 describes the testprocedure for the verification of downstream bi/gi values
for ADSL2plus operating mode.
© The Broadband Forum. All rights reserved.
135
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-11/TR-100: Verification of downstream bi/gi values
Test Configuration
Method of
Procedure
Expected Result
The ATU-C and ATU-R are connected in turn through the following
loops:
[1] ETSI loop #1, 250 m, -140 dBm/Hz AWGN, testprofile
B2P_RA_F_30000k
[2] ETSI loop #1, 1750 m, -140 dBm/Hz AWGN, testprofile
B2P_RA_I_30000k
[3] ETSI loop #1, 250 m, noise FB as defined in section B.1.3.3,
testprofile B2P_RA_I_30000k
[4] ETSI loop #1, 1750 m, noise FB as defined in section B.1.3.3,
testprofile B2P_RA_F_30000k
[1] Select test loop 1 from the set above.
[2] Train the modem in the test profile defined for the selected loop.
[3] From the DSLAM read the bi and gi values requested during
initialization
[4] Repeat steps 1 to 3 for other test loops in the set.
The RMSGIds shall be calculated in linear scale with gi values as
contained in the BiGi table and then rounded to nearest 100th of a dB.
The bi and gi values rounded to the nearest 100th of a dB in the bitsand-gains table shall comply with the following requirements
(Sections 8.5.2 and 8.6.4 of G.992.3[1])
[1] All bi values shall be in the [0 to BIMAXds] (bits) range, where 8
≤ BIMAXds ≤ 15.
[2] If bi > 0, then gi shall be in the [–14.54 to +2.50 + EXTGIds] (dB)
range where 0 ≤ EXTGIds ≤ MAXNOMPSDds – NOMPSDds.
[3] RMSGIds ≤ EXTGIds + 1/16th of a dB
[4] If bi > 0, then gi shall be in the [RMSGIds – 2.60 to RMSGIds +
2.60] (dB) range.
[5] If bi = 0, then gi shall be equal to 0 (linear) or in the [–14.54 to
RMSGIds + 0.10] (dB) range.
B.3.4 Loop Tests with Ports Set for Adaptive Rate
B.3.4.1 White Noise Impairment Only
Set noise generator G4 to apply white noise disturber at both ends of the total loop at –
140 dBm/Hz.
18 individual tests – 16 tests must be passed
© The Broadband Forum. All rights reserved.
136
ADSL2/ADSL2plus Performance Test Plan
TR-100
0
500
1000
1500
2000
2500
3000
3500
4000
Pass/Fail
Measured
Expected
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Loop Length [m,
loop#1]
B2P_RA_F_30000k
Upstream
Downstream
Sync Rate (kbps)
Sync Rate (kbps)
Noise Margin,
Reported (dB)
Table B.3-12/TR-100: White noise impairment, testprofile B2P_RA_F_30000k
24000
24000
22712
18112
12496
7472
4000
1648
504
1096
1096
1096
1096
1048
960
744
544
344
18 individual tests – 16 tests must be passed
© The Broadband Forum. All rights reserved.
Noise Margin,
Reported (dB)
7640
7640
7640
7640
7640
7640
4400
2000
792
Pass/Fail
19000
19000
18488
16280
12328
7640
4400
2000
792
Measured
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Expected
Class B
888
888
888
888
888
888
776
584
384
B2P_RA_I_30000k
Downstream
Sync Rate (kbps)
Class A
0
500
1000
1500
2000
2500
3000
3500
4000
Upstream
Sync Rate (kbps)
Expected
Loop Length [m, loop#1]
Table B.3-13/TR-100: White noise impairment, testprofile B2P_RA_I_30000k
137
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.3.4.2 Noise FB impairment
Set noise generators G1, G2, G4 and G8 for model FB impairment as defined in section
B.1.3
14 individual tests – 12 must be passed
Table B.3-14/TR-100: Noise FB ADSL2plus impairment, testprofile
B2P_RA_F_30000k
B2P_RA_F_30000
750
1000
13000
1250
816
11208
1750
600
7856
2500
280
2760
2750
184
1520
Noise Margin,
Reported (dB)
13000
Pass/Fail
1152
Measured
250
Expected
13000
Noise Margin,
Reported (dB)
1208
Pass/Fail
100
Measured
Expected
Downstream
Sync Rate (kbps)
Loop Length [m,
loop#1]
Upstream
Sync Rate (kbps)
14 individual tests – 12 must be passed
© The Broadband Forum. All rights reserved.
138
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-15/TR-100: Noise FB ADSL2plus impairment, testprofile
B2P_RA_I_30000k
Noise Margin,
Reported (dB)
7640
7640
7640
7640
7640
3408
1056
Pass/Fail
13000
13000
13000
12136
9088
3408
1056
Measured
Expected
Class B
888
888
888
832
640
328
128
Noise Margin,
Reported (dB)
Pass/Fail
Measured
Downstream
Sync Rate (kbps)
Class A
100
250
750
1250
1750
2500
3000
Expected
Loop Length [m, loop#1]
B2P_RA_I_30000
Upstream
Sync Rate (kbps)
B.3.5 Loop Tests with Ports Set For Fixed Rate
B.3.5.1 White Noise Impairment.
Set noise generator G4 to apply white noise disturber at both ends of the total loop at –
140 dBm/Hz.
a) Rate Profile: B2P_Fix_F_864k and B2P_Fix_I_864
12 individual tests – 12 shall be passed
© The Broadband Forum. All rights reserved.
139
ADSL2/ADSL2plus Performance Test Plan
TR-100
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
B2P_Fix_I_864k
Modem
Trained
(Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
B2P_Fix_F_864k
Modem
Trained
(Y/N)?
loop length (m,
loop#1)
Table B.3-16/TR-100: White noise fixed rate profile 864k
0
500
1000
2000
3000
Max. loop length requirement for
fast mode, test not to perform
3700
3800
Max. loop length requirement for
interleaved mode, test not to perform
b) Rate Profile: BP2_Fix_F_3456k and B2P_Fix_I_3456k
12 individual tests – 12 shall be passed
Table B.3-17/TR-100: White noise fixed rate profile 3456k
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
B2P_Fix_I_3456k
Modem
Trained (Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream
Noise Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_3456k
0
500
1000
2000
2750
Max. loop length requirement
for fast mode, test not to perform
3000
3100
Max. loop length requirement for
interleaved mode, test not to perform
© The Broadband Forum. All rights reserved.
140
ADSL2/ADSL2plus Performance Test Plan
TR-100
c) Rate Profile: B2P_Fix_F_7288k and BP2_Fix_I_7288k
10 individual tests – 10 shall be passed
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_I_7288k
Modem Trained
(Y/N)?
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_F_7288k
Modem Trained
(Y/N)?
loop length (m, loop#1)
Table B.3-18/TR-100: White noise fixed rate profile 7288k
0
500
1000
2000
Max. loop length requirement for
low delay, test not to perform
2400
2500
Max. loop length requirement for
high delay, test not to perform
d) Rate Profile: B2P_Fix_F_10000k and BP2_Fix_I_10000k
10 individual tests – 10 shall be passed
© The Broadband Forum. All rights reserved.
141
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-19/TR-100: White noise fixed rate profile 10000k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
B2P_Fix_I_10000k (Class A
only)
Modem Trained
(Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_10000k
0
500
1000
1750
Max. loop length requirement for
low delay, test not to perform
2200
2300
Max. loop length requirement for high
delay, test not to perform
B.3.5.2 European noise FB Impairments
Set noise generators G1, G2, G4 and G8 for model FB impairment, same pair ISDN and
white noise, as defined in section B.1.3.
a) Rate Profile: B2P_Fix_F_864k and B2P_Fix_I_864k
12 individual tests – 12 shall be passed
© The Broadband Forum. All rights reserved.
142
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-20/TR-100: Noise FB impairment fixed rate profile 864k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
B2P_Fix_I_864k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_864k
0
500
1000
1750
2500
Max. loop length requirement for fast
mode, test not to perform
2700
2800
Max. loop length requirement
for interleaved mode, test not to
perform
b) Rate Profile: B2P_Fix_F_3456k and BP2_Fix_I_3456k
10 individual tests – 10 shall be passed
Table B.3-21/TR-100: Noise FB impairment fixed rate profile 3456k
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_I_3456k
Modem Trained
(Y/N)?
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin, Reported
(dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_3456k
0
500
1000
1750
Max. loop length requirement for
fast mode, test not to perform
2050
2150
Max. loop length requirement for
interleaved mode, test not to
perform
© The Broadband Forum. All rights reserved.
143
ADSL2/ADSL2plus Performance Test Plan
TR-100
c) Rate Profile: B2P_Fix_F_7288k and B2P_Fix_I_7288k
8 individual tests – 8 shall be passed
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_I_7288k
Modem Trained
(Y/N)?
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_F_7288k
Modem Trained
(Y/N)?
loop length (m, loop#1)
Table B.3-22/TR-100: Noise FB impairment fixed rate profile 7288k
0
500
1000
Max. loop length requirement for
fast mode, test not to perform
1200
1300
Max. loop length requirement for
interleaved mode, test not to
perform
d) Rate Profile: B2P_Fix_F_10000k and BP2_Fix_I_10000k
8 individual tests – 8 shall be passed
© The Broadband Forum. All rights reserved.
144
ADSL2/ADSL2plus Performance Test Plan
TR-100
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_I_10000k (Class A
only)
Modem Trained
(Y/N)?
Downstream Noise
Margin, Reported
(dB)
Upstream Noise
Margin, Reported
(dB)
B2P_Fix_F_10000k
Modem Trained
(Y/N)?
loop length (m, loop#1)
Table B.3.23/TR-100: Noise FB impairment fixed rate profile 10000k
0
500
750
Max. loop length requirement for
low delay, test not to perform
1150
1250
Max. loop length requirement for high
delay, test not to perform
B.3.5.3 Target Noise margin consideration
For a detailed description of this test see section B.2.5.3.
a) Target noise margin consideration upstream
© The Broadband Forum. All rights reserved.
145
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-24/TR-100: Target noise margin consideration upstream
[1] Rate parameters acc. B2P_Fix_F_3456k.
[2] Set noise generators for model FB impairment as defined in
section B1.3.1.
Method
of [1] Start test at maximum length from test B.2.5.2 b) at which
the modem was able to train successfully
Procedure
[2] Allow to train for 60 seconds to achieve showtime.
[3] Record the upstream and downstream reported noise
margin in the table below. (Read the margin one minute
after show time is reached.)
[4] If margin upstream > target noise margin + 1 dB, increase
the loop length by:
next loop length step =
floor[(reported
noise
margin
–
target
noise
margin)]/0.6*50m
[5] Repeat steps 2 – 3, if the modem was able to train increase
the loop length 50 meters and follow steps 5, until the
modem fails to connect.
[6] After the first length for which the modem fails to connect
attempt the next incremented loop length. (50 meters
greater then the failing loop length of step 5.) If the CPE
connects record the reported upstream and downstream
margin.
Expected Result to If the modem reaches showtime for a certain loop length, then
the related reported noise margin must meet the pass/fail
Pass
criteria as stated in section B.1.4.
Test Configuration
Table B.3-25/TR-100: Target noise margin consideration upstream - Results table
Pass / Fail
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin,
Reported (dB)
Modem
Trained (Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_3456k
note 1
..
1950
200
2050
…
…
Note 1: max. length from test B.3.5.2 b), at which the modem was able to train
successfully
b) Target noise margin consideration downstream
© The Broadband Forum. All rights reserved.
146
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-26/TR-100: Target noise margin consideration downstream
[1] Rate parameters acc. B2P_Fix_F_3456k.
[2] Set noise generators G4 to apply white noise impairment as
defined in section B1.3.
Method
of [1] Start test at maximum length from test B.2.5.1 b) at which
the modem was able to train successfully
Procedure
[2] Allow to train for 60 seconds to achieve showtime.
[3] Record the upstream and downstream reported noise
margin in the table below. (Read the margin one minute
after show time is reached.)
[4] If margin downstream > target noise margin + 1 dB,
increase the loop length by: next loop length step =
floor[(reported
noise
margin
–
target
noise
margin)/1.0]*50m
[5] Repeat steps 2 – 3, if the modem was able to train increase
the loop length 50 meters and follow steps 5, until the
modem fails to connect.
[6] After the first length for which the modem fails to connect
attempt the next incremented loop length. (50 meters
greater then the failing loop length of step 5.) If the CPE
connects record the reported upstream and downstream
margin.
Expected Result to If the modem reaches showtime for a certain loop length, then
the related reported noise margin must meet the pass/fail
Pass
criteria for the rate adaptive tests (10% limit) as stated in
section B.1.4.
Test Configuration
© The Broadband Forum. All rights reserved.
147
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-27/TR-100: Target noise margin consideration downstream - Results
table
Pass / Fail
Downstream
Noise Margin,
Reported (dB)
Upstream Noise
Margin, Reported
(dB)
Modem Trained
(Y/N)?
loop length (m,
loop#1)
B2P_Fix_F_3456k
note 2
..
2900
2950
3000
…
…
Note 2: max. length from test B.3.5.1 b), at which the modem was able to train
successfully
B.3.6 Performance in L2 mode
For a detailed description of this test see section B.2.6.
© The Broadband Forum. All rights reserved.
148
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-28/TR-100: ADSL2plus performance in L2 mode
Test Configuration
For this test the following parameters have to be configured in addition
to the defined parameters in B1.3:
[1] Test configuration: see figures 4 to 6, depending on customer
interface of modem.
[2] Configure the line with testprofile B2P_RA_I_30000k (note 1)
[3] PMMode: Low power state allowed (bit 0 = 0, bit 1 = 1)
[4] PM-Parameters:
i. Minimum (Net) Data rate DS in Low Power state(L2Min):
128 kbps
ii. L0-TIME:
60 sec.
iii. L2-TIME:
15 sec.
iv. L2-ATPR: 3 dB
v. L2-ATPRT: 18 dB
[5] Set loop simulator to: 1500 m with ETSI loop #1.
[6] Set noise generators for model ADSL2plus FB impairment as
defined in section B1.3.2.
Method
of [1] Traffic generator/analyzer switched off or 120 kbps DS traffic
[2] Synchronize the line with the required parameters and note the
Procedure
actual downstream sync rate and noise margin (first L0 noise
margin) value.
[3] Wait for automatic transition to L2 with power trims. If automatic
transition does not occur within 120 seconds, force L2 entry via
the G997.1[4] PMFS MIB element.
[4] After 5 minutes read the actual downstream noise margin value
(L2 noise margin) and DS Tx power (ACTATPds).
[5] Switch on traffic generator with 90% of DS sync rate, to exit from
L2 state to L0 state.
[6] Wait 60 seconds and read the actual DS noise margin (second L0
noise margin) value.
Expected Result to [1] first L0 noise margin >= 5.0 dB.
[2] L2 noise margin >= 5.0 dB.
Pass
[3] Second L0 noise margin >= first L0 noise margin –1 dB.
[4] No CRC error during the entire test.
Note 1: The B2P_RA_I_30000k minimum net data rate in the L2 mode test
shall be equal to L2_min.
B.3.7 Performance with configured RFIBANDds
With this test the performance of an ADSL2 DSLAM – Modem combination with
configured RFIBANDSds shall be proven. The following two RFI Bands shall be
considered:
• RFI Band 1:
1.800 up to 2000 kHz / -80 dBm/Hz
• RFI Band 2:
2.173,5 up to 2.190,5 kHz / -80 dBm/Hz
© The Broadband Forum. All rights reserved.
149
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-29/TR-100: ADSL2plus performance with configured RFIBANDds
For this test the following parameters have to be configured in
addition to the defined parameters in B1.2:
[1] Testprofile: B2P_RA_F_30000k.
[2] RFIBANDds and PSDMASKds parameters:
• breakpoints representing RFI Band 1: subcarrier/level:
t(2)= 417, PSD(2)= -80 dBm/Hz
t(3)=464, PSD(3)= -80 dBm/Hz
• breakpoints representing RFI Band 2: subcarrier/level:
t(6)= 504, PSD(6)= -80 dBm/Hz
t(7)= 508, PSD(7)= -80 dBm/Hz
[3] Set noise generators for model ADSL2plus FB impairment
as defined in section B1.3.2.
Method
of Steps per loop length:
[1] Synchronize the line with the required parameters.
Procedure
[2] Note the actual downstream bitrate.
[3] After 60 seconds note the reported DS noise margin.
Expected Result to [1] fast and stable synchronization.
Pass
[2] DS Bitrate: see table B.3-30.
[3] noise margin >= 5.0 dB.
Test Configuration
Note: acc. G.992.5[2] and G.997.1[4] each RFI band [ti, ti+1] is also
represented in the PSDMASKds with 4 breakpoints (t(i + 1), PSD(i + 1)) to
(t(i + 4), PSD(i + 4)). The related breakpoints for PSDMASKds using the
steepest allowed slope are:
•
•
for RFI band 1: t(1) = 394 / -47 dBm/Hz, t(4) = 484 / -50 dBm/Hz
for RFI band 2: t(5) = 484 / -50 dBm/Hz, t(8) = 511 / -75,5 dBm/Hz
6 individual tests – 5 must be passed
Table B.3-30/TR-100: performance requirements for Tests B.3.7
Downstream
Noise
measured
Margin,
Reported
(dB)
DS sync rate [kbps]
loop length
[m, loop#1]
expected
100
11000
250
11000
750
10904
1250
9704
2000
5904
2750
904
© The Broadband Forum. All rights reserved.
Pass / Fail
150
ADSL2/ADSL2plus Performance Test Plan
TR-100
B.3.8 ADSL2plus European tests operating with DS-PSD mask above ADSL
With this test the performance of an ADSL2 DSLAM – Modem combination with DSPSD mask above ADSL with additional activated RFI Bands acc. B.3.7 shall be proven.
Note : the handshake carrier set B43c shall be used to initiate startup.
Table B.3-31/TR-100: ADSL2plus performance with DS-PSD mask above ADSL
For this test the following parameters have to be configured in
addition to the defined parameters in B1.2:
[1] Testprofile: B2P_RA_F_30000k
[2] RFIBANDds and PSDMASKds parameters:
ƒ breakpoints DS-PSD mask above ADSL:
t(1)= 233 (1004,8125 kHz), PSD1= -95 dBm/Hz
t(2)= 260 (1121.25 kHz), PSD2= -36.5 dBm/Hz
Ö t(2) as close as possible to tone 256
ƒ breakpoints representing RFI Band 1: subcarrier/level:
t(4)= 417, PSD(4)= -80 dBm/Hz
t(5)=464, PSD(5)= -80 dBm/Hz
ƒ breakpoints representing RFI Band 2: subcarrier/level:
t(8)= 504, PSD(8)= -80 dBm/Hz,
t(9)= 508, PSD(9)= -80 dBm/Hz,
[3] Set noise generators for model ADSL2plus FB impairment as
defined in section B1.3.2.
Method
of Steps per loop length:
[1] Synchronize the line with the required parameters.
Procedure
[2] Note the actual downstream bitrate.
[3] After 60 seconds note the reported DS noise margin.
Expected Result to [1] Fast and stable synchronization
[2] DS Bitrate: see table B.3-32
Pass
[3] Noise margin >= 5.0 dB
Test Configuration
Note: acc. G.992.5[2] and G.997.1[4] each RFI band [ti, ti+1] is also
represented in the PSDMASKds with 4 breakpoints (t(i + 1), PSD(i + 1)) to
(t(i + 4), PSD(i + 4)). The related breakpoints for PSDMASKds using the
steepest allowed slope are:
•
•
for RFI band 1: t(3) = 394 / -47 dBm/Hz, t(6) = 484 / -50 dBm/Hz
for RFI band 2: t(7) = 484 / -50 dBm/Hz, t(10) = 511 / -75,5 dBm/Hz
5 individual tests – 4 must be passed
© The Broadband Forum. All rights reserved.
151
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table B.3-32/TR-100: performance requirements for Tests B.3.8
Downstream
Noise
measured
Margin,
Reported
(dB)
DS sync rate [kbps]
loop length
[m, loop#1]
expected
100
5344
250
5184
750
4600
1500
3520
2250
1016
© The Broadband Forum. All rights reserved.
Pass / Fail
152
ADSL2/ADSL2plus Performance Test Plan
TR-100
Annex C Physical Layer Test Cases for systems using
G.992.3/5 annex C
For further study.
© The Broadband Forum. All rights reserved.
153
ADSL2/ADSL2plus Performance Test Plan
TR-100
Annex D ADSL2plus Equivalent disturber frequency domain
profiles of generators G1 and G2 for the European test sets.
Linear interpolation of the PSD in dBm/Hz against log(f) shall be used to calculate the
values between breakpoints.
D.1
Equivalent disturber frequency domain profiles for Annex A.2 test set
Table D.1-1/TR-100: NT-profiles for Annex A
X.NT.FA
[kHz]
[dBm/Hz]
1
-20
15
-20
24
-20.9
30
-21
45
-23
60
-24.7
138
-24.9
151
-28
207
-28.7
300
-30.3
358
-32.8
407
-36.7
500
-48.6
594
-62.3
755
-62.3
1059
-73.7
1221
-75.5
1400
-77.9
1750
-85
2212
-92
2800
-96
5274
-97.1
30000
-97.1
X.NT.FB
[kHz]
[dBm/Hz]
1
-25.8
15
-25.6
24
-26.5
30
-26.8
61
-30.5
138
-30.8
149
-33
200
-33.5
308
-35.2
375
-38.5
456
-46.9
605
-68.4
755
-68.4
980
-77.3
1128
-80.8
1402
-83.7
1750
-90.8
2208
-97.6
2800
-101.4
5274
-102.2
30000
-102.2
© The Broadband Forum. All rights reserved.
X.NT.FD
[kHz]
[dBm/Hz]
1
-90.9
3.99
-90.9
4
-85.9
25.875
-27.9
138
-27.9
229.6
-82.8
686
-89.9
1411
-89.9
1630
-99.9
5275
-101.9
12000
-101.9
30000
-101.9
154
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table D.1-2/TR-100: LT-profiles for Annex A
X.LT.FA
[kHz]
[dBm/Hz]
1
-20.1
15
-20
30
-21.6
45
-24.1
60
-26.9
68
-27.6
137.99
-27.9
138
-26.1
300
-27
470
-27.8
1106
-27.8
1622
-37.8
2208
-39.1
2500
-50.7
3001.5
-71.3
3175
-87.7
3750
-87.7
4000
-91
4300
-94.3
4545
-96.2
7225
-97.1
30000
-97.1
X.LT.FB
[kHz]
[dBm/Hz]
1
-25.7
15
-25.6
30
-27.1
45
-29.6
65
-32.6
137.99
-32.8
138
-31.7
272
-32.6
414
-34.2
1103
-34.2
1622
-44.2
2208
-45.5
2500
-57.1
3001.5
-77.7
3175
-94.1
3750
-94.1
4000
-97.2
4300
-100.2
4545
-101.6
7225
-102.2
30000
-102.2
© The Broadband Forum. All rights reserved.
X.LT.FD
[kHz]
[dBm/Hz]
1
-90.9
3.99
-90.9
4
-85.9
80
-65.9
137.99
-37.6
138
-29.9
1104
-29.9
1622
-39.9
2208
-41.2
2500
-52.8
3001.5
-73.4
3175
-89.9
3750
-89.9
4545
-99.9
7225
-101.9
12000
-101.9
30000
-101.9
155
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table D.1-3/TR-100 19 ADSL2plus downstream disturber noise models
Freq
[kHz]
0
4
4
80
138
138
1104
1622
2208
2500
3001.5
3175
3750
4545
7225
12000
Template
X.LT.FD19
ADSL2+ PSD [dBm/Hz]
[dBm/Hz]
-101
-93.3
-101
-93.3
-96
-88.3
-76
-68.3
-47.7
-40
-40
-32.3
-40
-32.3
-50
-42.3
-51.3
-43.6
-62.9
-55.2
-83.5
-75.8
-100
-92.3
-100
-92.3
-110
-102.3
-112
-104.3
-112
-104.3
Table D.1-4/TR-100 19 ADSL2plus upstream disturber noise model
Freq
[kHz]
0
4
4
25.875
138
229.6
686
1411
1630
5275
12000
Template
X.NT.FD19
ADSL2plus PSD [dBm/Hz]
[dBm/Hz]
-101
-93.3
-101
-93.3
-96
-88.3
-38
-30.3
-38
-30.3
-92.9
-85.2
-100
-92.3
-100
-92.3
-110
-102.3
-112
-104.3
-112
-104.3
© The Broadband Forum. All rights reserved.
156
ADSL2/ADSL2plus Performance Test Plan
D.2
TR-100
Equivalent disturber frequency domain profiles for Annex A.3 test set
Table D.2-1/TR-100 19 ADSL2plus over ADSL downstream PSD and crosstalk
noise model for CAL = 12
Freq
[kHz]
50
80
138
168
372
677
1108
1622
2208
3200
15000
CAL=12
PSD
X.LT.C12
[dBm/Hz] [dBm/Hz]
-90.5
-82.8
-83.8
-76.1
-41.4
-33.7
-53.0
-45.3
-53.0
-45.3
-56.7
-49.0
-61.3
-53.6
-76.1
-68.4
-82.5
-74.8
-110.0
-102.3
-110.0
-102.3
Table D.2-2/TR-100 19 ADSL2plus over ADSL downstream PSD and crosstalk
noise model for CAL = 36
Freq
[kHz]
50
80
138
200
300
301
500
782
1011
1131
1132
2208
4600
15000
CAL=36
PSD
X.LT.C36
[dBm/Hz] [dBm/Hz]
-90.5
-82.8
-83.8
-76.1
-41.4
-33.7
-63.7
-56.0
-72.5
-64.8
-76.0
-68.3
-83.8
-76.1
-93.5
-85.8
-93.5
-85.8
-83.5
-75.8
-53.0
-45.3
-53.0
-45.3
-110.0
-102.3
-110.0
-102.3
© The Broadband Forum. All rights reserved.
157
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table D.2-3/TR-100 19ADSL2plus over ADSL downstream PSD and crosstalk noise
model for CAL = 52
CAL=52
Freq
PSD
X.LT.C52
[kHz] [dBm/Hz] [dBm/Hz]
50
-90.5
-82.8
80
-83.8
-76.1
138
-41.4
-33.7
200
-63.7
-56.0
300
-72.5
-64.8
301
-82.5
-74.8
452
-91.3
-83.6
551
-83.5
-75.8
552
-53.0
-45.3
2208
-53.0
-45.3
4600
-110.0
-102.3
15000
-110.0
-102.3
D.3
Equivalent disturber frequency domain profiles for Annex B.3 test set
Table D.3-1/TR-100: NT-profiles for Annex B
X.NT.FA
[kHz]
[dBm/Hz]
1
-20.1
15
-20
30
-21.5
45
-24.1
69
-27.6
108
-27.6
119
-24.8
276
-25.3
310
-30.6
400
-36.2
595
-62.4
755
-62.4
1082
-74.2
1183
-76.3
1400
-77.9
1750
-85
2212
-92
2800
-96
5274
-97.1
30000
-97.1
X.NT.FB
[kHz]
[dBm/Hz]
1
-25.8
15
-25.6
30
-27.2
45
-29.6
70
-32.6
108
-32.7
119
-30.7
276
-31.4
310
-35.5
400
-41.1
605
-68.4
755
-68.4
1082
-80.2
1186
-82.2
1400
-83.7
1750
-90.8
2208
-97.6
2800
-101.4
5274
-102.2
30000
-102.2
© The Broadband Forum. All rights reserved.
X.NT.FD
[kHz]
[dBm/Hz]
1
-79.9
50
-79.9
80
-75.2
120
-27.9
276
-27.9
491
-87.7
686
-89.9
1411
-89.9
1630
-99.9
5275
-101.9
12000
-101.9
30000
-101.9
158
ADSL2/ADSL2plus Performance Test Plan
TR-100
Table D.3-2/TR-100: LT-profiles for Annex B
X.LT.FA
[kHz]
[dBm/Hz]
1
-20.1
15
-20
30
-21.5
44
-24.1
60
-26.9
68
-27.6
99
-27.6
200
-28.7
254
-29.5
255
-26.7
414
-27.8
1104
-27.8
1622
-37.8
2208
-39.1
2500
-50.7
3000
-71.3
3174
-87.8
3750
-87.8
4550
-96.2
7200
-97.1
30000
-97.1
X.LT.FB
[kHz]
[dBm/Hz]
1
-25.7
15
-25.6
30
-27.1
45
-29.6
59
-32
66
-32.5
99
-32.6
200
-33.6
253
-34.3
254
-32.5
414
-34.2
1103
-34.2
1622
-44.2
2208
-45.5
2500
-57.1
3001.5
-77.7
3175
-94.1
3750
-94.1
4220
-99.5
4550
-101.6
7200
-102.2
30000
-102.2
© The Broadband Forum. All rights reserved.
X.LT.FD
[kHz]
[dBm/Hz]
1
-79.9
93.1
-79.9
209
-55.4
253.99
-41.9
254
-29.9
1104
-29.9
1622
-39.9
2208
-41.2
2500
-52.8
3001.5
-73.4
3175
-89.9
3750
-89.9
4545
-99.9
7225
-101.9
12000
-101.9
30000
-101.9
159
ADSL2/ADSL2plus Performance Test Plan
Appendix I : Effect of
Manufacturing (Informative)
TR-100
Statistical
Variability
in
CPE
This section gives information on the effect of statistical variability in CPE
manufacturing variation on the statistics involved with the methodology, how the
“expected” datarate requirements were determined for section A.1 “North American Test
Set”.
Methodology
This section describes the essentials of the methodology, how the “expected” datarate
requirements were determined for section A.1 “North American Test Set”. Only the
essentials needed for understanding the statistics in the next section are given. Full details
of the testmethodology can be found in TR-067[19] Annex Y “Test Event Methodology”.
A number of CPE’s have been submitted by CPE vendors to an test event, open to service
providers and vendors. It is assumed that these modems represent randomly selected
samples from a production lot.
Based on the methodology in Annex Y, a selection of 2 candidate modem-types was
established, which correspond with the best 2 modem-types that are fully standard
compliant. Preliminary datarates where measured, and that modem-type was selected
with the lowest datarate. This was done for each test point, and for upstream and
downstream direction independently.
In a further refinement process, 4 units of the selected modem-type were measured and
datarates recorded. The final datarate requirements have been established, by taking
picking the rates of 4 units, and determining picking the lowest of these 4 rates.
Statistical Analysis
In this section, the analysis is given of the statistics of the rate-requirement (Rr) which is
determined by taking the lowest rates out of N rates (Ri: I=1…N) obtained from
different units. It can be assumed that the manufacturing variance on equipment
corresponds with a gaussian distribution of the rates R.
The probability density function (PDF) of the rate-requirement is given by below
equation:
Prob(Rr) = Prob(R1=Rr) * Prob(R1 < All other Ri) +
Prob(R2=Rr) * Prob(R2 < All other Ri) +
…
Prob(RN=Rr) * Prob(RN < All other Ri) +
Therefore:
PDF ( Rr ) = N * PDF _ gaussian ( Rr ) * (CDF _ gaussian ( Rr )) ( N −1)
© The Broadband Forum. All rights reserved.
160
ADSL2/ADSL2plus Performance Test Plan
TR-100
This is illustrated in Figure 1. In this figure the distributions are plotted in function of the
“normalised R”:
Normalised_R = (R-R_mean)/R_sigma
With
R_mean = mean value of rates R- distribution
R_sigma = variance of rates R-distribution.
The solid curve gives the gaussian PDF of the individual rates Ri. The dashed curve
gives the PDF of the rate-requirement Rr for N=4 as in [Annex Y].
Figure W.1 PDF of the individual rates Ri, and the Rate-requirement Rr.
The overlapping part in Figure 1 correspond with modem-units for which Ri < Rr, i.e. the
part of the production lot which could show a fail on the performance tests, although the
modem-type is compliant to this TR.
The probability that this occurs can be calculated as :
+∞
Failprobability =
∫
Pr ob( Rr ) * Pr ob( Ri < Rr ) * dRr
−∞
This probability has been evaluated, with results showing that the large majority, (but not
all) of randomly-selected units will pass the rate requirements.
© The Broadband Forum. All rights reserved.
161
ADSL2/ADSL2plus Performance Test Plan
TR-100
Appendix II : Computing the Maximum Number of Frames Per
Second (Informative)
The following logic was used to compute the Maximum number of Frames per Second
(MAX FPS), testsection 8.1, given a data rate at the ATM layer, and a frame size.
The RFC-2684 encapsulation (LLC bridging) process adds 10 bytes per frame. The
AAL5 CPCS-PDU includes an 8-byte trailer. Therefore, the number of cells required, for
each frame, is as follows:
⎛ frame size + 18 ⎞
ATM cells per frame = ceil ⎜
⎟
48
⎠
⎝
In the above, the frame size is in bytes, and the ceil function rounds up to the next integer
value.
Each ATM cell is composed of 53 bytes, so that
ATM layer bits per frame = 53 × 8 × ATM cells per frame
Finally, since the data rate is that rate available to the ATM layer, the Maximum number
of Frames per Second (Max FPS) is:
⎛
⎞
data rate
⎟⎟ ,
Max FPS = floor ⎜⎜
⎝ ATM layer bits per frame ⎠
In the above, the data rate is in bits per second, and the floor function rounds down to the
next integer value.
The calculated framerate in section 8 of this document is based on a frame size without
preservation of Ethernet FCS.
END OF DOCUMENT
© The Broadband Forum. All rights reserved.
162
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising