Measuring and Qualifying Upstream Signals

1
Measuring and Qualifying Upstream Signals

• SCTE
• 9/12/07

2
My Business Card

• Tom Scanlin
• Sales and Support Engineer
• Sunrise Telecom Broadband
• 708-751-7510
• 800-297-9726 tech support
• tscanlin_at_sunrisetelecom.com
• www.sunrisetelecom.com

3
Purpose

• Better understand how to make upstream signals
  and measurements
• What are the signal impairments on the reverse
  path

4
Agenda

• DOCSIS path
• 16 QAM Advantages and Challenges
• Measurements and Impairments
• Case Study

5
History of DOCSIS

• DOCSIS 1.0
• Open standard for high-speed data over cable
• Best-effort
• 1st products certified 1999
• DOCSIS 1.1
• Quality-of-Service (QoS) service flows
• BPI with Certificates
• Improved privacy with key distribution
  encryption processes
• SNMP for network management security

6
History of DOCSIS (Cont.)

• DOCSIS 2.0
• Goal greater throughput robustness on Return
  Channel
• Adds 64 128 QAM modulation to Return Channels
• Higher symbol rate up to 5.12 Msps (BW 6.4)
• Adds Forward Error correction, Trellis coding
  programmable interleaving to Return channel
• Adds multiple modulation access schemes
• DOCSIS 3.0
• Channel bonding (Increase capacity)
• Enhanced network security
• Expanded addressability (IPv6)

7
Basic DOCSIS Setup
8
Why 16-QAM?

• Higher upstream data throughput required for
• Voice
• Peer to Peer

9
Upstream 16 QAM Challenges

• Once interference occurs in voice the data cannot
  be retransmitted.
• Measurements are more difficult because the
  signals are bursty.
• 16 QAM looses 3 dB of headroom because the
  maximum modem output is 55 dBmV as opposed to 58
  dBmV for QPSK.

10
More Upstream Challenges with 16 QAM

• 16 QAM is less robust than QPSK
• Requires better SNR and MER
• QAM means that the carrier is amplitude modulated
  and therefore more susceptible to amplitude based
  impairments such as
• Ingress
• Micro-reflections
• Compression

11
What needs to be done before launching 16 QAM?

• Making 16-QAM work reliably requires attention to
  several details
• The cable modem termination system (CMTS)
  configuration must be optimized for 16-QAM
• Its also important to configure the CMTS
  parameter called modulation profile correctly.
• The entire cable system needs to be
  DOCSIS-compliant
• Return and Forward Sweep!
• Leakage kept to levels below that required by the
  FCC
• If it leaks out, it leaks in!

12
Recommended Network Specifications

• Part 76 of the FCC Rules
• DOCSIS for upstream and downstream
• NCTA Recommended Practices for upstream carriers

13
Upstream Signal Measurements
14
You cant get there from here
The problem could be here
To CMTS Receive Port
Spare Splitter Leg
Optical Receiver
Fiber Node
Optical Receiver
or here
Optical Receiver
or here
CoaxDist.Network
The actual Call might be here
or the problem could be anywhere in these three
nodes.
15
Characterizing the Upstream
Return Path Verification, Test Troubleshooting
Test signal injected in field measured on
analyzer Measure MER, BER, Constellation, Freq.
Response, Group Delay
16
Spectrum Analyzer Upstream Measurements

• Upstream Carrier Levels
• Spectrum Analysis
• Constellation Measurements and Diagnosis
• MER, BER, and Constellation Analysis
• Upstream Linear Distortion Measurements

17
Upstream Level MeasurementThe First Step

• Verify the upstream carrier amplitude at the
  input to the CMTS upstream port is within spec.
• Usually 0 dBmV at the input, some systems may
  vary.
• Can be measured using peak power on the preamble
  of the carrier
• An average power measurement could also be made
  on a constant carrier injected at the correct
  level.
• Be careful of mixed modulation profile
  measurements remember there is a 3 dB difference
  between QPSK and 16 QAM
• Measure total power at the input to the CMTS
  (lt35dBmV, TP)

18
Max Hold

• Using Max Hold will allow you to get a relative
  reading on the Cable Modems in the return.
• This Method is not very accurate, but does
  provide a good approximation.

19
Upstream Level Measurement

• Measurement made in the zero span mode
• Peak power of the preamble

20
Spectrum Analysis -CNR -C/I
21
DOCSIS 1.1 Upstream RF Channel Transmission
Characteristics
22
Upstream CNR

• Check the upstream carrier-to-noise,
  carrier-to-ingress, and carrier-to-interference
  ratios
• DOCSIS assumes a minimum of 25 dB for all three
  parameters
• This is measured at the CMTS upstream port
• Remember that we loose 3 dB of dynamic range with
  16 QAM
• CNR and SNR are different measurements!
• The correct noise power bandwidth is equal to the
  symbol rate of the upstream carrier

23
CNR or SNR

• CNR is a measurement performed on RF signals
• Raw carrier power to raw noise power in the RF
  transport path only
• Ideal for characterizing network impairments
• SNR is a pre-modulation or post-detection
  measurement performed on baseband signals
• Includes noise in original signal, transmitter or
  modulator, transport path, and receiver
  demodulator
• Ideal for characterizing end-to-end
  performancethe overall signal quality seen by
  the end user

24
CMTS Upstream SNR Measurement

• Broadcom burst demodulator chips used in a CMTS
  provides an upstream SNR estimate.
• Other factors may degrade CMTS-reported SNR, even
  when CNR is good including improper modulation
  profiles, bad timing errors, and poor headend
  combiner/splitter isolation. These of course
  would be system return path problems.
• Impulse noise and certain other fast transient
  impairments generally will not show up in CMTS
  SNR estimate.
• CMTS-reported SNR will always be less thanor at
  best equal toCNR, but should never be better
  than CNR.

25
Upstream Spectrum Analysis

• Make sure noise floor of system is being
  displayed 10 db out of the spectrum analyzer
  noise floor
• Use peak hold to capture transients

26
Different BW Different Power
1.5 MHz QPSK
6 MHz QAM

• The higher the symbol rate of a digital carrier,
  the greater the bandwidth
• The area under the curve represents the power of
  the signal.

27
CNR using a Noise Marker
28
Upstream Carrier-to-Interference
29
Common Path Distortion

• A corroded connection causes mixing
• The resulting impedance mismatch also causes
  reflections
• The mixing products are reflected right back into
  the return amplifier.
• The diplex filter takes out everything above 42
  MHz.

Corroded Connection
27
Downstream Signals
Difference frequencies reflected upstream
(6, 12, 18, 24)
30
CPD in 6 MHz Intervals

• Because the channels in the forward system are 6
  MHz apart, the sum difference frequencies occur
  at 6 MHz intervals as well.

31
Constellation Analysis

• Patterns in the Constellation

32
MER, A Better Measurement

• A better parameter than SNR is modulation error
  ratio (MER) or error vector magnitude (EVM)
• MER takes into account
• CNR
• Phase Noise (jitter of phase of QAM modulators
  carrier)
• Intermod Distortions
• Compression of Lasers and Amplifiers
• Frequency Response
• THE SUM OF ALL EVILS
• MER is a single figure of merit for the quality
  of an RF QAM modulated signal.
• MER and EVM are the same thing. MER is expressed
  in dB EVM is expressed in .
• A direct measurement of the digital signals
  modulation quality
• Can be directly linked to BER

33
Vectors and 16 QAM
Q 90
11
1011
1011
1111
10
1010
1110
I 180
I 0
00
11
10
01
01
0101
0001
00
0000
0100
Q 270
34
Vectors and 16 QAM
Q 90
11
1011
1011
1111
10
1010
1110
I 180
I 0
00
11
10
01
01
0101
0001
00
0000
0100
Q 270
35
Modulation Error Ratio

• MER is defined as follows
• MER is expressed in dB.

RMS Error Magnitude
Ideal Symbol
Average Symbol Magnitude
36
Modulation Error Ratio

• Minimum recommended downstream MER (includes 3 to
  4 dB of headroom for reliable operation)
• 64-QAM 27 dB
• 256-QAM 31 dB
• QPSK 13 dB
• 16 QAM 20 dB

Graphic courtesy of Sunrise Telecom
37
Introduction to BER

• Bit Error Rate (BER) is an important concept to
  understand in any digital transmission system
  since it is a major indicator of the quality of
  the digital system.
• As data is transmitted some of the bits may not
  be reproduced at the receiver correctly. The
  more bits that are incorrect, the more the signal
  will be affected.
• BER is a ratio of incorrect bits to the total
  number of bits measured.
• Its important to know what portion of the bits
  are in error so you can determine how much margin
  the system has before failure.

38
What is BER?

• BER is defined as the ratio of the number of
  wrong bits over the number of total bits.
• BER is measured by sending a known string of bits
  and then counting the errored bits vs. the total
  number of bits sent.

Sent Bits 1101101101 Received Bits 1100101101
error
1
of Wrong Bits
0.1


BER
of Total Bits
10
39
Scientific Notation Question
Question Scientific Notation is used on test
equipment to confuse technicians and make them
think the equipment is smarter than they are.
True or False ?
40
Scientific Notation Answer

• FALSE Scientific Notation is used to display
  very large numbers in a limited amount of space.

41
Scientific Notation
1 Error in 10 bits 1/10
0.1 1.0 E-1 1 Error in 100
bits 1/100
0.01 1.0 E-2 1 Error in 1000 1/1000
0.001 1.0 E-3 1 Error
in 1 million 1/1,000,000 0.0000001
1.0 E-6 1 Error in 1 billion 1/1,000,000,000
0.000000001 1.0 E-9 2 Errors in 1 billion
2/1,000,000,000 0.000000002 2.0 E-9
42
What is BER?

• BER is normally displayed in Scientific Notation.
• The more negative the exponent, the better the
  BER.
• Better than 1.0E-6 is needed after the FEC for
  the system to operate.

Lower and Better BER
43
Forward Error Correction

• In every MPEG digital receiver, there is a FEC
  decoder that can actually repair damaged data.
• Forward error correction (FEC) is a digital
  transmission system that sends redundant
  information along with the payload, so that the
  receiver can repair the damaged data and
  eliminate the need to retransmit.
• It only works if the BER is higher than 1E-6
  before the FEC decoder

44
Pre and Post FEC BER

• To get an accurate idea of the BER performance
  you need to know both the pre and post FEC bit
  error rate.
• The FEC decoder needs a BER of better than 1 E-6
  in order to operate.
• Post FEC Bit errors are not acceptable.
• You should look at both the Pre and Post FEC BER
  to determine if the FEC is working to correct
  errors and if so how hard.

Pre FEC BER
Post FEC BER
FEC Decoder
45
Typical BER/MER Results
64 QAM 256 QAM BER MER
MER Quality 0E-0 gt35
gt35 Excellent 1E-8 27-34
31-34 Good 1E-6 23-26
28-30 Marginal 1E-5 lt23
lt28 Fail
Note Set-top boxes can tolerate some Post FEC
errors, but Cable Modems cannot.
46
Noise and Intermittents

• Errors caused by noise or intermittent causes can
  have the same BER, but very different effects.
• Errors that are spread out are due to noise
  problems
• Errors that are grouped are due to intermittent
  problems such as ingress or loose connectors.

Spaced Errors 1101101011010011100 Burst
Errors 1111101011101101101
This Example Shows the Same Error Rate But the
Burst Errors are More Difficult to Correct
47
Constellation Analysis
48
A Good 16 QAM Constellation
49
CPD and Noise
50
Laser Clipping
51
Ingress
52
A Good 16 QAM Constellation
Zero Bit Errors
53
Adaptive Equalizers

• Corrects for Frequency Response imperfections
• Corrects for Group Delay
• Show impedance mismatches

54
Adaptive Equalizers
55
Microreflections

• Micro-reflections are impedance mismatches
• In the real world of cable networks, 75 O
  impedance is at best considered nominal
• Micro-reflections cause group delay and frequency
  response problems.
• Impedance mismatches are everywhere connectors,
  amplifiers inputs and outputs, passive device
  inputs and outputs, and even the cable itself
• Upstream cable attenuation is lower than
  downstream cable attenuation, so upstream
  micro-reflections tend to be worse.
• Anywhere an impedance mismatch exists, some of
  the incident energy is reflected back toward the
  source

56
Micro-reflections

• Higher orders of modulation are affected by
  micro-reflections to a much greater degree so 16
  QAM is affected more than QPSK
• Upstream micro-reflections and group delay are
  minimized by using adaptive equalizers. This
  feature is available in DOCSIS 1.1 and 2.0 CMTSs
  , but not 1.0.

57
Microreflections
Causes

• Damaged or missing end-of-line terminators
• Damaged or missing chassis terminators on
  directional coupler, splitter, or multiple-output
  amplifier unused ports
• Loose center conductor seizure screws
• Unused tap ports not terminatedthis is
  especially critical on low value taps
• Unused drop passive ports not terminated
• Use of so-called self-terminating taps at feeder
  ends-of-line

58
Microreflections
Causes (contd)

• Kinked or damaged cable (includes cracked cable,
  which causes a reflection and ingress)
• Defective or damaged actives or passives
  (water-damaged, water-filled, cold solder joint,
  corrosion, loose circuit board screws, etc.)
• Cable-ready TVs and VCRs connected directly to
  the drop (return loss on most cable-ready devices
  is poor)
• Some traps and filters have been found to have
  poor return loss in the upstream, especially
  those used for data-only service

59
Microreflections
-10 dBc _at_ lt0.5 µsec -20 dBc _at_ lt1.0 µsec -30 dBc
_at_ gt1.0 µsec
60
Microreflections

• Heres an example An approx. -33 dBc echo at
  just over 1 µsec
• This echo meets the DOCSIS upstream -30 dBc at
  gt1.0 µsec parameter however this is sufficient to
  cause some amplitude and group delay ripple

61
Equalizer Display

• Equalizer presentation shows how hard the
  equalizer circuit is working to counteract the
  effects of reflections and linear distortions
• Marker for Distance Calculation

Distance 180 88 m
62
Amplitude Ripple ( Frequency Response)
63
Amplitude Ripple
An in-service spectrum analyzer measurement
64
Group Delay

• Different signals travel through the same medium
  at different speeds. This is Group Delay
• Group delay is defined in units of time,
  typically nanoseconds (ns) over frequency. In
  other words how much GD per each MHz.
• In a system, network or component with no group
  delay, all frequencies are transmitted through
  the system, network or component with equal time
  delay
• Frequency response problems in a CATV network
  will cause group delay problems
• Group delay is worse near band edges and diplex
  filter roll-off areas

65
Upstream frequency

• Keep the 16-QAM digitally modulated carrier well
  away from diplex filter roll-off areas (typically
  above about 3538 MHz), where group delay can be
  a major problem
• Choose an operating frequency that will minimize
  the likelihood of group delay
• Frequencies in the 2035 MHz range generally work
  well
• Group delay may still be a problem when the
  frequency response is flat

66
Group Delay
67
Group Delay Measurement
68
Some things to check out!

• Before adding a 16 QAM carrier the following
  should be checked
• Compression of the return laser due to added
  carrier or a carrier with added bandwidth
• MER and BER over a period of time
• Group Delay of a new carrier
• MER and BER of the new carrier.
• Amplitude Ripple
• Microreflections

69
Upstream Spectrum Display Showing Compression
70
16 QAM Constellation with Clipping
71
Statistics Mode
72
Group Delay Measurement
73
Frequency Response of an Upstream Carrier
74
A Case Study
75
Upstream Spectrum
76
The Constellation
77
Oops!
78
Amplitude Ripple
D 492(VP/F) 492(87/.4MHz) 1100 feet
79
Group Delay
80
Bit Errors
81
Moral of the Story?

• CNR and Distortion measurements from a spectrum
  analyzer are great but, dont tell the whole
  story.
• Other digital measurements are advised using a
  vector analyzer to ensure 16 QAM reliability
• MER and BER
• Group Delay and other Equalizer measurements
• Constellation
• Statistic Measurement

82
16-QAM Pre-Launch Checklist

• CMTS modulation profile optimized for 16-QAM
• Vector Analysis, not just spectrum analysis
• Entire cable networkheadend, distribution
  network and subscriber dropsDOCSIS-compliant
• Select upstream frequency that avoids diplex
  filter roll-off area
• Forward and reverse properly aligned
• Signal leakage and ingress management
• Good installation practices

83
Measurement Summary

• Check for laser clipping
• Measure over time
• Measure for frequency response of the carrier
• Measure group delay of the carrier
• Measure MER and BER of upstream carrier
• Can be accomplished by inserting a 16 QAM carrier
  at the EOL and using a digital analyzer in the
  headend.

84
References
Ron Hranac wrote the book

• Hranac R., CNR versus SNR March 2003
  Communications Technology
• Hranac R., Spectrum analyzer CNR versus CMTS
  SNR September 2003 Communications Technology
• Hranac R., 16 QAM Plant Preparation
• Hranac R., Deploying VOIP on the Outside Plant
• Hranac R., Linear Distortions, Last 2 issues
  of CT Magazine

85
References

• RF Impairments in the Return Path and their
  impact on DOCSIS performance, by Jack Moran,
  Motorola
• National Cable Television Associations
  Recommended Practices for Measurements on Cable
  Television Systems, 2nd Edition, October 1997
  Supplement on Upstream Transport Issues.
• Broadband Return Systems for HFC Cable TV
  Networks, by Donald Raskin and Dean Stoneback
• Return Path Level Selection, Set Up, and
  Alignment Procedure, Motorola 1997
• Modern Cable Television Technology, by Walter
  Cicora, James Farmer and David Large

86
More References

• Mystified by Return Path Activation? Get your
  Upstream Fiber Links Aligned, by Ron Hranac,
  Communications Technology, March 2000
• Seek Balance in All Things A Look at Unity
  Gain in the Upstream Coax Plant, by Ron Hranac,
  Communications Technology, June 2000
• A Primer on Common Path Distortion, by Nick
  Romanick, Communications Technology, April 2001

87
Thank You!