1. Suzanne Ewert Systems Engineer
DOCSIS 3.0 Overview
Suzanne Ewert
Systems Engineer

2. Agenda Evolution of DOCSIS Motivation - Why DOCSIS 3.0?
DOCSIS 3.0 Features Overview
Downstream Bonding Details
Upstream Bonding Details
DOCSIS 3.0 and M-CMTS Comparisons
Migration Strategy
Cisco VDOC
Cisco Architecture for D3.0 & M-CMTS
Summary

3. Evolution of DOCSIS

4. Evolution of DOCSIS Pre-DOCSIS DOCSIS 1.0 DOCSIS 1.1
MSO’s needed a service offering for the residential market
Consumer demands dictated the need for something faster than dial-up
Proprietary and expensive
DOCSIS 1.0
MSO’s needed a standardized solution (i.e. cheaper)
Consumer demands dictated the need for additional bandwidth
Competing against DSL
DOCSIS 1.1
MSO’s needed a way to protect their infrastructure and offer differentiated services
MSO’s needed to expand, start targeting the commercial market
Competing against DSL, ISDN, and T1
Standard defined:
security between the CMTS and CM (BPI+)
extensive QOS functionality
38Mbps x 9Mbps service offering

5. Evolution of DOCSIS (cont)
MSO’s needed a way to offer a synchronous service
VoIP and business services
Consumer demands dictated the need for more upstream bandwidth
Gaming
Consumer owned servers (Peer-to-Peer)
Standard defined:
Expanded upstream channel widths to include 6.4MHz
Expanded upstream modulation schemes to include 32QAM, 64QAM, and 128QAM
S-CDMA
38Mbps x 27Mbps service offering

6. Motivation - Why DOCSIS 3.0?

7. Business Drivers for D3.0 More HD Video Services More SD Video Content
Competition against FTTH - Deliver 100 Mbps
Broadband Internet Services Growth
Migration from Web to Web2.0, Video Streaming, P2P TV
Increased per home consumption
IP Video over DOCSIS(VDOC)
High definition Video to multiple devices
PCs, Hybrid STBs, portable devices
Migration from Broadcast to Unicast services (VoD, Startover)
Commercial services
High BW data services
High BW Ethernet/L2VPN service
Video conferencing
Despite the improvements that have occurred as DOCSIS has evolved, maximum data rates to and from cable modems are pretty much topped out. Competition and the desire to provide new services are driving the need for even greater throughput in our DOCSIS networks.
We're limited by the fact that the maximum raw data rate to or from cable modems is ultimately constrained by what a single 6 MHz wide channel can carry in the downstream, or what a single 6.4 MHz wide channel can carry in the upstream.
Enter DOCSIS 3.0 and something called channel bonding
Video in the Internet is driving data throughput at 41% compound annual growth rate.
By 2012, 50% of internet traffic is expected to be video
More HD Video Services
Growth plans to 100+ HD channels
More SD Video Content
Expansion to nx100 SD chs to compete w/ satellite
Personalized Video Services
Migration from Broadcast to Unicast services
VoD, Startover, MyPrimetime, etc

8. Next Generation Connected Home
Stored music
In any room
Internet
Next Gen
MR-DVR
Internet video
On HDTV
Photos
From PC
Multi-Media
Client Gateway
Multi-Media
Client Gateway
No New Wires Technology
Outside
The
Home
Next Gen
MR-DVR
DVR content
Over the Internet
Network
IP Service
Gateway
Photos
From PC
Ethernet
Internet video
On HDTV
DVR content
Over the Internet PC
Stored music
In any room
Multi-Media
Service Gateway

9. Spectral Reclamation Solutions
SDV – Switched Digital Video
Node splits
Narrowcast QAM injection
Analog reclamation
Use every channel available
1 GHz upgrade
MPEG-4
SDV Offer more HD and SD content using less total RF spectrum with the same STB
Only transmit the content being actively watched
Could make more QAMs available for DOCSIS and VOD if QAM sharing is implemented
Node splits Physically reduce the homes passed per HFC node, thus reduce contention per home for Unicast services
Decombine more attractive
Triggers additional QAMs and CMTS Ports
Broadcast to narrowcast QAM injection Reduce broadcast domains to smaller DOCSIS & video service groups
Ultimately a complete Unicast lineup on a per node basis
Analog reclamation for more digital spectrum More QAM channels for Digital Broadcast, VoD, SDV and DOCSIS
Use every channel available Manage the channel lineup, fill in the gaps, mitigate noise to enable all spectrum
1GHz upgrade Make new spectrum for new CPE above 860 MHz

10. Overall Industry Objectives
DOCSIS 3.0
M-CMTS
Goal:
More aggregate speed
More per-CM speed
Enable New Services
Components:
Channel Bonding
IPv6
Multicast
AES
Goal:
Increase Scalability
Reduce Cost
Components:
Low Cost E-QAM
CMTS Core Processing
Better stat muxing with bigger “pipe”
Offer >37 Mbps for single CM

11. DOCSIS 3.0 Features Overview

12. DOCSIS 3.0 Features MAC Layer Network Management Network Layer
Downstream Channel Bonding
Upstream Channel Bonding
Network Layer
IPv6 support
IP Multicast (IGMPv3/MLDv2, SSM, QoS)
Security
Certificate Revocation Management
Runtime SW / Config validation
Enhanced Traffic Encryption (AES)
Certificate Convergence
Early Authentication & Encryption
TFTP Proxy
Network Management
Diagnostic Log (Flaplist)
Extension of Internet Protocol Data Records (IPDR) usage
Capacity management
Enhanced signal quality monitoring
Physical Layer
Switchable 5-42 MHz, 5-65 MHz, or 5-85 MHz US band
S-CDMA active code selection with new Logical channel
Commercial Services
T1/E1 Circuit Emulation support
Stress IP Multicast

13. DOCSIS 3.0 Features – Physical Layer CMTS Deployment Models
Integrated CMTS
Implements the network ports and RF interface ports in a single network element
Modular CMTS
Implements the network ports and URFI ports in a modular core network element and the DRFI ports in a external EQAM
A DEPI tunnel is used to encapsulates the downstream channels from the M-CMTS core to the EQAM
A DTI server is used to synchronize the M-CMTS core and all associated EQAM’s

14. DOCSIS 3.0 Features – MAC Layer
Downstream Channel Bonding
Allows a CM to receive data on multiple receive channels using a single service flow
At least 4 channels must be used to equal 150+ Mbps
Upstream Channel Bonding
Allows a CM to transmit data on multiple transmit channels using a single service flow
At least 4 channels must be used to equal 100+ Mbps

15. DOCSIS 3.0 Features – Network Layer
IPv6 support
Built in support for IPv6
Modems can be provisioned using IPv4, IPv6, or both
Provides transparent IPv6 connectivity to CPE’s
IP multicast support
Supports delivery of source specific multicast (SSM) streams to CPE’s
CMTS controlled layer-2 multicast forwarding mechanism
Introduces “group service flow” concept to provide QOS to multicast streams

16. DOCSIS 3.0 Features – Security
CMTS to CM Privacy Features
128-bit AES traffic encryption (performed in hardware)
Early CM authentication and traffic encryption (EAE)
MMH (Multilinear Modular Hash) algorithm for CMTS MIC (message integrity check)
Prevent Unauthorized Access
Enhanced secure provisioning features
Source IP address verification (SAV)
TFTP proxy and configuration file learning;
Certificate Revocation
Encryption support for new method of multicast messaging.
EAE = Early Authentication Encryption

17. DOCSIS 3.0 Features – Network Management (cont)
Security Management
IETF deprecated the previous NmAccess approach
In order to address the new D3.0 features and the IETF’s decision:
Extensions were built to report configuration status, error conditions and statistics of the new security features
Replacement of NmAccess is required using a method compatible with the SNMPv3 framework
Accounting Management
SNMPv3 polling/trapping
IPDR (IP Detail Record) support is expanded to include the new D3.0 features
New D3.0 features primarily include channel bonding, M-CMTS, IPv6, and IP Multicast
- IPDR defines a way to stream the export of accounting records with less resource taxing. IRDP was supported in the D2.0 OSSI spec, but expanded upon.

18. CableLabs DOCSIS 3.0 Qualification Tiers
Bronze
DS channel bonding
IPv6 CM provisioning without dual stack, basic IPv6 forwarding for CPE
Basic DOCSIS 2.0 multicast features, IPv6 multicast support for CM provisioning
No US channel bonding, No S-CDMA, No AES
Silver
Bronze features plus:
US channel bonding
Additional IPv6 support
AES, SSM, Bonded multicast, S-CDMA w/o bonding, parts of IPDR
Gold
Full DOCSIS 3.0 support

19. DOCSIS 3.0 Downstream Channel Bonding Details

20. Downstream Bonding - Features
Packet bonding of a minimum of 4 channels
Delivers in excess of 150 Mbps and 50 Mbps US
Non-disruptive technology
Seamless migration from DOCSIS 1.x/2.0
M-CMTS and high density I-CMTS cards
EQAMs
New hardware required for scalability and cost reduction
New CM silicon required

21. Channel Bonding
In a nutshell, channel bonding means data is transmitted to or from CMs using multiple individual RF channels instead of just one channel
Channels aren't physically bonded into a gigantic digitally modulated signal; bonding is logical
With DOCSIS 3.0, data is transmitted to modems using multiple channels
With DOCSIS 1.x & 2.0, data is transmitted to modems using one channel
Let's say you want to increase the downstream data rate between the CMTS and modems from today's single 6 MHz wide channel limit of Mbps.
􀂃If you were to spread your downstream data payload across four 6 MHz wide channels, the combined raw data rate using 256-QAM on each channel would be Mbps x 4 = Mbps.
􀂃A DOCSIS 3.0 modem incorporates a special tuner capable of simultaneously receiving data from those four channels. To the modem, the four channels are the logical equivalent of one large bonded channel, even though we're using four physically separate channels. They don't even have to be adjacent channels!
Want more? Bonding, say, 10 channels, will yield Mbps x 10 = Mbps, and bonding 24 channels works out to 24 x Mbps = 1, Mbps, or just over 1 Gbps. Yikes!
􀂃The same channel bonding concept is applicable to the upstream, giving us the ability to go far beyond DOCSIS 2.0's per-channel limit of Mbps. How does 4 x Mbps = Mbps—or more—sound?

22. DOCSIS 3. 0 Downstream Channel Bonding with Today’s DOCSIS 2
DOCSIS 3.0 Downstream Channel Bonding with Today’s DOCSIS 2.0 Deployments
Universal Edge QAM
Wideband MAC
Traditional Cable Modems
WCM
Wideband Downstream
D3.0 CM
Docsis 3.0 Bi-Dir CM CM
Traditional DOCSIS
Wideband Bearer Channel
The relationship of Wideband and traditional QAM modulators is shown in the figure above. Traditional DOCSIS CMs share a common downstream and are spread across several upstreams. The WCM uses the same traditional infrastructure and adds more downstream or upstream capacity. The multiple QAM carriers within a Wideband channel are coupled together to form what looks like a multi-carrier PHY. The coupling does not actually occur in the PHY; it occurs within the transmission convergence layer between the MAC and PHY. This is very significant because it implies that large pipes – upwards of a Gigabit – can be constructed out of today’s inexpensive QAM technology. The Wideband Protocol will support any number of QAM carriers, although the implementations of the WCMTS and WCM will set operational limits. 22

23. DOCSIS 3.0 Registration Diagram
D3.0 CM acquires QAM/FEC lock of DOCSIS DS channel
SYNC, UCD, MAP messages
D3.0 CM performs usual US channel selection, but does not start initial ranging
MDD message
D3.0 CM performs bonded service group selection, and indicates via initial ranging
B-INIT-RNG-REQ message
D3.0 CM transitions to ranging station maintenance as usual
Usual DOCSIS initial ranging sequence
DHCP DISCOVER packet
DHCP OFFER packet
DHCP REQUEST packet
DHCP RESPONSE packet
TOD Request/Response messages
TFTP Request/Response messages
D3.0 CM provides Rx-Chan(s)-Prof
REG-REQ message
REG-RSP message
D3.0 CM receives Rx-Chan(s)-Config
D3.0 CM confirms all Rx Channels
REG-ACK message
Usual BPI init. If configured

24. Reasons DRFI went Beyond D2.0 RFI
Applies to CMTS, D3.0, or multi-carrier CMTS DS connector
Cleaned up ambiguity in 2.0 and lower
Noise dBmV changed to dBc
Allows more channels per connector
DOCSIS 2.0 and lower was only single carrier
M-CMTS architecture & D3.0 both reference DRFI
Less expensive E-QAMs, MxN mac domains
Performance goal was analog protection given analog ch lineup of 2-13 ( MHz)
Digital chs justified to upper end of spectrum
Criteria was 60 dB CNR for all combined sources
Not necessary for digital communication nor sparser lineup
denotes the logarithmic power ratio relative to the strongest carrier in the channel
block. The out-of-band spurious emissions requirements assume a test condition with a contiguous block of N
combined channels commanded to the same power level, and for this test condition "dBc" should be interpreted as
the average channel power, averaged over the block, to mitigate the variation (see Table A–3) in channel power
across the block, which is allowed with all channels commanded to the same power.

25. Single Carrier DRFI Annex A & B Channel BW 8 & 6 MHz
Variable depth interleaver
HRC, IRC
64 & 256 QAM
dBmV
N=1 : 60 1
Harmonic Related Carrier
Incremental Related Carrier
Center Frequency
Must 91 <-> 867 MHz
May 57 <-> 999 MHz

26. Single Carrier DRFI (cont)
In-band spurious, distortion & noise MER
Unequalized MER > 35 dB
Equalized MER > 43 dB
In-band spurious & noise ≤ -48 dBc
Spurious & noise within ±50 kHz of carrier excluded
Phase noise (single carrier)
kHz: -33 dBc
kHz: -51 dBc
50 kHz - 3 MHz: -51 dBc
Double sided noise power

27. Single Carrier DRFI (cont)
Output return loss
>14 dB within active output ch from MHz
>13 dB within active output ch from MHz
>12 dB in every inactive ch from MHz
>10 dB in every inactive ch from MHz
Power accuracy per channel +/- 2 dB
RF muting ≥73 dB below aggregate power

28. Power Output for Multiple Carriers per RF Spigot
dBmV 1
60-ceil[3.6*log2(N)] dBmV 1 2 1 2 3 1 2 3 4
RF muting ≥73 dB below aggregate power N
dBmV 1 60 2 56 3 54 4 52 8 49 16 45 32 42
Why is it done like this?
Multiple chs create more pwr & distortions
Attempt to keep constant wattage output
DS laser concerns (Pwr/Hz)

29. Step Size ≤ 0.2 dB (configuration granularity)
Multi-Carrier DRFI
In-band spurious & noise ≤ -48 dBc
Spurious & noise within ±50 kHz of carrier is excluded
When N > 1, noise outside Nyquist BW is excluded
Phase noise (multi-carrier)
kHz: -33 dBc
kHz: -51 dBc
Double sided noise power
dBmV
Step Size ≤ 0.2 dB (configuration granularity)
Power Diff (adjacent ch) ≤ 0.5 dB
Power Diff (non-adj ch) ≤ 1.0 dB 1 2 3 4

30. Out-of-Band Noise and Spurious Emissions
Edge to
750 kHz
750 kHz to
6 MHz
6 to 12 MHz
12 to 18 MHz
Other Chs (47 MHz – 1 GHz)
2nd & 3rd
Harmonics
N=1
< -58
< -60
< -65
< -73
Greater of
-63 dBc or
*log(n) dBc
N=2
< -64
< -70
N=3
< -63.5
< -67
< -68
N=4
< -63

31. DOCSIS 3.0 DS Considerations
Frequency Assignments
CMTS may be limited to 860 MHz or 1 GHz
CM’s may be limited to 50 or 60 MHz passband
Testing and maintaining multiple DS channels
Physical channels have not changed for DOCSIS 3.0
Test equip with built-in CM’s need to support bonding
DS isolation issues
DS channel bonding max power with 4 freqs stacked
Four channels stacked on 1 connector limited to 52 dBmV/ch
DOCSIS 1.x/2.0 DS is 61 dBmV max output

32. DOCSIS 3.0 – Upstream Channel Bonding Details

33. Upstream Bonding Service Drivers
Competition against FTTH
Deliver 20+ Mbps
High BW residential data
User generated content
Video and photo uploads
Proliferation of social sites
Video conferencing
TelePresence
Commercial service
High BW symmetrical data services
Bonded T1
High BW Ethernet/L2VPN service

34. Upstream Bonding - Features
Packet Striping of a minimum of 4 channels
Delivers in excess of 50 Mbps
AES and scalability require hardware upgrade
New CM silicon required
Phased and seamless technology migration

35. Upstream Channel Bonding
Upstream bonding
Single flow can consume all BW on multiple USs
Continuous Concatenation & Fragmentation (CCF)
Improved form of concatenation and fragmentation that is needed for DOCSIS 3.0 operation
The CM has a buffer with a 1000 bytes to send. Its requests for 1000 on US1,
the CM receives grants on US1, US2 and US3. The size of the grant may depend
on the load on that channel
The CM does not send packets since the grant might not align with packet
Boundaries
Instead the CM sends “segments”
Each segment has a sequence number, and a pointer field so that individual packets
Can be extracted (the pointer field is similar to the one used on the MPEG pointer
In the DS

36. D2.0 is Still Not Used 27.2 Mbps total aggregate speed
Achieved 18 Mbps for single CM on US
Fragmentation and concatenation with a huge max burst
Linerate possible of ~ 27 Mbps
Make sure 1.0 CMs, which can’t fragment, have a max burst < 2000 B
2.0 increases the EQ tap length from 8 to 24
Supported in ATDMA & mixed mode
Off by default
There are other factors that can directly affect performance of your cable network such as the QoS Profile, noise, rate-limiting, node combining, over-utilization, etc. Most of these are discussed in detail in:
Knowing what throughput to expect is the first step in determining what subscribers' data speed and performance will be. Once it is determined what is theoretically possible, a network can then be designed and managed to meet the dynamically changing requirements of a cable system.
Reject(na) indicates a reject nack. This can occur when a 1.0 modem doesn’t respond correctly to the DOCSIS mode of the US port it is connected to.
Right now, load balancing is not supported with DOCSIS 2.0 settings because of load balance weights. Weights are related to the aggregate speed of the “pipe”. In a mixed (DOCSIS 1.x and 2.0) environment, the 1.x CMs could have a weight of and the 2.0 CMs could have a weight of 15 Mbps.
Symbol Rate, ksym/sec
Channel Bandwidth, MHz
QPSK Raw Data Rate, Mbps
QPSK Nominal Data Rate, Mbps
QAM-16 Raw Data Rate, Mbps
QAM-16 Nominal Data Rate, Mbps
QAM-64 Raw Data Rate, Mbps
QAM-64 Nominal Data Rate, Mbps
1280
1.6
2.56
2.3
5.12
4.6
7.68
6.9
2560
3.2
10.24
9.2
15.36
13.8
5120
6.4
20.48
18.4
30.72
27.5

37. Upstream Adaptive Equalization Example
Upstream 6.4 MHz bandwidth 64-QAM signal
Before adaptive equalization:
Substantial in-channel tilt caused correctable FEC errors to increment at a rate of about 7000 errored codewords per second (232 bytes per codeword). The CMTS’s reported upstream MER (SNR) was 23 dB.
After adaptive equalization:
DOCSIS 2.0’s 24-tap adaptive equalization—actually pre-equalization in the modem—was able to compensate for nearly all of the in-channel tilt (with no change in digital channel power). The result: No correctable or uncorrectable FEC errors and the CMTS’s reported upstream MER (SNR) increased to ~36 dB.
This slide illustrates an example of upstream pre-equalization in action. To demonstrate the performance of DOCSIS 2.0’s 24-tap pre-equalization, a cable modem was set up to transmit a 6.4 MHz-wide 64-QAM signal at a center frequency of 48 MHz. This frequency forced the signal through the rolloff area of the cable modem’s internal low pass filter, causing substantial in-channel tilt. The upper photo shows the 64-QAM signal as received by the CMTS without adaptive pre-equalization, and the lower photo shows the same signal at the CMTS input after the modem’s adaptive pre-equalization was turned on.
Before adaptive equalization: The in-channel tilt caused correctable FEC errors to increment at a rate of about 7000 errored codewords per second (232 bytes per codeword). The CMTS’s reported upstream unequalized MER (“SNR”) was 23 dB.
After adaptive equalization: The modem’s 24-tap pre-equalizer was able to compensate for nearly all of the in-channel tilt (with no change in digital channel power). The result: No correctable or uncorrectable FEC errors and the CMTS’s reported upstream unequalized MER increased to ~36 dB.

38. DOCSIS 3.0 Upstream Channel Bonding
Bonding process is controlled by the CMTS
Bandwidth grants are given per flow across one or more upstream channels as CM’s make requests
New packet streaming protocol called Continuous Concatenation and Fragmentation.
Allows a looser coupling between requests and grants
Enables the CM to have multiple requests outstanding simultaneously
Bonding Mechanism
Upstream channels are synchronized to a master clock source

39. DOCSIS 3.0 US Considerations
Frequency Stacking Levels
What is the CM max output with multiple channels stacked
Could it cause laser clipping?
Diplex Filter Expansion to 85 MHz
If amplifier upgrades are planned for 1 GHz, then pluggable diplex filters may be warranted to expand to 85 MHz on the US…one truck roll
Still must address existing CPE equipment in the field and potential overload
Monitoring, Testing, & Troubleshooting
Test equipment needs to have D3.0 capabilities
As explained for DS considerations, any new technology will have some trade-offs and potential pitfalls that we must understand and plan for in advance. The following will discuss some of these issues:
Why it’s Needed – This can range from competitive pressure, to higher tiers of service, to more customers signing up.
Frequency Stacking Levels & Placement – What is the modem maximum US output with four channels stacked and do the channels have to be contiguous?
Isolation Concerns – Whenever applications have different service groups, we have overlaid networks. Signals destined for one node could “bleed” over to another.
US Frequency Expansion to 85 MHz – Amplifier upgrades are occurring now. It’s best to make the truck roll once. Think about diplex filters, line EQs, step attenuators, taps, etc.

40. DOCSIS 3.0 Upstream Input Spec
To address this potential issue where a modem today transmits near max power of 54 dBmV for 64-QAM, the specification has changed the CMTS US port level setting to allow it to be 6 dB lower. This means the CMTS can be set lower so modems can be placed on those high value taps without changing HE or plant losses. This is at the expense of lower MER/SNR readings.
The lowest setting on the CMTS today is -1 dBmV for 6.4 MHz wide channel. The range allowed on the CMTS is dictated by DOCSIS 2.0 and lower and says -1 to + 29 dBmV for 6.4 MHz and related to channel width, also known as symbol rate or baud.
D3.0 identified this potential issue and forced D3.0 CM vendors to support a transmit of 3 dB higher than the 2.0 spec. So 64-QAM has to at least max out at 57 dBmV with a single channel.
To keep the CM inexpensive, constant wattage device, etc, the max output will be dictated by how many US frequencies are active on the port for bonding. Four channels stacked will be 57-6 = 51 dBmV per ch. So, the overall effect from 2.0 to 3.0 is really 3 dB difference. I can see why the spec lowered the nominal setting allowed on the CMTS so cable operators didn’t need to lower tap faceplates or drop padding on every US port on the CMTS to keep D3.0 CMs online if they max out.
Most systems will leave the default of 0 dBmV and adjust padding appropriately. Cisco has a command to keep CMs online that are maxed out, power-adjust continue 4 (default). Many customers set it to 6 to give some more room to work with.

41. DOCSIS 3.0 US Considerations (cont) US Frequency and Level Issues
Max Tx for D QAM for 1 channel is 54 dBmV
D3.0 US channel max power
Tx for D3.0 TDMA
dBmV (32 & 64-QAM)
58 dBmV (8 & 16-QAM)
61 dBmV (QPSK)
Tx for D3.0 S-CDMA
dBmV (all modulations)
Max Tx per channel for 4 freqs stacked at 64-QAM ATDMA is only 51 dBmV & 53 for S-CDMA
As with DS issues, there are also US issues that need to be addressed. US bonding has not been pursued at this point because most people haven’t even exploited D2.0 US capabilities. This does not mean we should avoid the potential issues that will arise. Eventually, we will want to offer US speeds greater than what a single channel modem can offer of ~ 25 Mbps. This will require more US spectrum, D3.0 CMs, and CMTS linecards with US bonding capability.
Some of the potential issues are levels and frequency assignments. Activating multiple frequencies per US connector on a 3.0 CM has different max power per channel vs a D2.0 CM. Max transmit for a D2.0 CM using 64-QAM for 1 channel is 54 dBmV.
D3.0 US channel max power is: dBmV when using 32 & 64-QAM, 58 dBmV when using 8 & 16-QAM, & 61 dBmV when using QPSK. D3.0 S-CDMA US max power is: dBmV for all modulations.
As shown above, it can be seen that the max power for 1 channel on the connector has been raised by 3 dB over a D2.0 CM, but max transmit per channel for four frequencies stacked using 64-QAM ATDMA is only 51 dBmV & 53 for S-CDMA.
US passband has also changed in the D3.0 specification. Frequency assignments were 5 to 42, 55, & 65 MHz for Euro-DOCSIS, but it has been extended to 85 MHz. The option of going higher is good for future spectrum re-allocation and avoiding known bad frequencies on the US. Some things need to be considered though and that includes, diplex filters, line EQs, step attenuators, and CPE overload. If incorporating any of these devices in the plant, they may need to be replaced for the new frequency split. Also, can current customer premise equipment (CPE) like settops and TVs handle a potentially high level of “noise” from a modem at 50 MHz or higher?

42. DOCSIS 3.0 US Considerations (cont) US MER/SNR Issues
Increasing channel width from 3.2 to 6.4 keeps same average power for single carrier
SNR drops by 3 dB or more
Keeping same power/Hz could cause max Tx level from CM’s and/or laser clipping/overload
Equalized vs unequalized MER readings
Modulation profile choices
QPSK for maintenance, 64-QAM for Data, 16-QAM for VoIP?
Pre-EQ affect
Great feature in 1.1 & > CMs, but could mask issues
As explained for DS considerations, any new technology will have some trade-offs and potential pitfalls that we must understand and plan for in advance. The following will discuss some of these issues:
Why it’s Needed – This can range from competitive pressure, to higher tiers of service, to more customers signing up.
Frequency Stacking Levels & Placement – What is the modem maximum US output with four channels stacked and do the channels have to be contiguous?
Isolation Concerns – Whenever applications have different service groups, we have overlaid networks. Signals destined for one node could “bleed” over to another.
US Frequency Expansion to 85 MHz – Amplifier upgrades are occurring now. It’s best to make the truck roll once. Think about diplex filters, line EQs, step attenuators, taps, etc.

43. DOCSIS 3.0 US Considerations (cont) Channel Placement
Frequencies can be anywhere in US passband and do not need to be contiguous
It may be wise to keep relatively close so plant problems like attenuation and tilt don’t cause issues
CM should have some dynamic range to allow specific channels to be a few dB different vs. other channels
Channels are separate and can have different phy layer attributes such as modulation, channel width
Since each US channel used for bonding is an individual channel, frequencies can be anywhere in the US passband and do not need to be contiguous. Although, it may be wise to keep relatively close so plant problems like attenuation and tilt don’t cause issues. The CM should have some dynamic range to allow specific channels to be a few dB different vs. other channels. Since the transmitters (channels) are separate, they don't have to be contiguous and can have different physical layer attributes like; modulation, channel width, tdma or scdma, etc

44. ATDMA General Deployment Recommendations
After increasing CW to 6.4 MHz, measure & document unequalized US MER at multiple test points in the plant
Use PathTrak Return Path Monitoring System linecard
Or Sunrise Telecom Upstream Characterization toolkit
25 dB or higher Unequalized MER is recommended
Less than 25 dB reduces operating margin
Check US MER as well as per-CM MER
Pick freq < 30 MHz - away from diplex filter group delay
Make sure latest IOS version is running on CMTS
Turn on Pre-Equalization
Many times doubling the US channel width will indicate more issues with the plant than actually increasing the modulation to 64-QAM. Linear impairments like group delay and micro-reflection will not be apparent with a spectrum analyzer, but will severely degrade US modulation error ratio (MER). MER is also known as US SNR as listed on the CMTS. After increasing the channel width to 6.4 MHz, it’s imperative to measure and document unequalized US MER at multiple test points in the plant. Unequalized means per-CM US MER without pre-eq activated. Some tools that can be used include: JDSU PathTrak Return Path Monitoring System linecard or the Sunrise Telecom Upstream Characterization toolkit.
The recommended unequalized MER is 25 dB or higher. Less than 25 dB reduces operating margin. Be sure to check US MER as well as per-CM MER
If group delay is suspect in addition to long amplifier cascades, it may be necessary to pick a frequency below 30 MHz, away from diplex filter group delay.
If group delay is an issue and difficult to overcome with frequency placement, it may be possible to activate Pre-Equalization for per-CM low MER issues.
Be sure the latest IOS version is running on the CMTS and proper modulation profiles.

45. DOCSIS 3.0 and M-CMTS Comparisons

46. DOCSIS 3.0 Migration: M-CMTS
Current CMTS
DOCSIS 2.0 US
HFC
DS Bonding and Existing DOCSIS 1.x/2.0 CMs
Edge QAMs

47. M-CMTS Network Topology
Emphasize – can use Modular DS in non-bonded configuration. Also called Narrowcast

48. M-CMTS Key DOCSIS 3.0 enabling technology
DTI server = Symmetricom. External timing server.
Key DOCSIS 3.0 enabling technology
DS scalability of DOCSIS 1.x/2.0
Easy migration to DOCSIS 3.0 DS channel bonding
Enables service convergence and QAM sharing (Video and Data)

49. Supports DS Bonding and Existing DOCSIS 1.x/2.0 CMs
DOCSIS 3.0: M-CMTS
CMTS Core
DOCSIS 3.0 Bonded US
HFC
Supports DS Bonding and Existing DOCSIS 1.x/2.0 CMs
Edge QAMs

50. Supports DS Bonding and Existing DOCSIS 1.x/2.0 CMs
DOCSIS 3.0: I-CMTS
High Density Linecards
I-CMTS
DOCSIS 3.0 Bonded US
HFC
DOCSIS 3.0 Bonded DS
Supports DS Bonding and Existing DOCSIS 1.x/2.0 CMs

51. Spectrum Example

52. Bandwidth Management

53. Bandwidth Management Solutions
SDV
Offer more HD and SD content using less total RF spectrum with the same STB
Only transmit the content being actively watched
Could make more QAMs available for DOCSIS and VOD if QAM sharing is implemented
Node splits
Physically reduce the homes passed per HFC node, thus reduce contention per home for Unicast services
Decombine more attractive
Triggers additional QAMs and CMTS Ports

54. Bandwidth Management Solutions (cont)
Traffic “Grooming”
MPEG-4
Broadcast to narrowcast QAM injection
Reduce broadcast domains to smaller DOCSIS & video service groups
Ultimately a complete Unicast lineup on a per node basis
Analog reclamation for more digital spectrum
More QAM channels for Digital Broadcast, VoD, SDV and DOCSIS
Use every channel available
Manage the channel lineup, fill in the gaps, mitigate noise to enable all spectrum
1GHz upgrade
Make new spectrum for new CPE above 860 MHz

55. 1GHz Bandwidth Enhancement & Segmentation
1 GHz Upgrade
1GHz Bandwidth Enhancement & Segmentation
Network Impact
<= 750 MHz of BW may not be enough
Node splitting & SDV alone do not solve HFC BW problem
1 GHz BW upgrade required
1GHz Network Benefits
Value added capacity
60 analog 6 MHz chs gained
Minimal cost per home passed cost to implement
Electronic-only drop-ins in most cases
1 GHz is a cost-effective tool to increase broadcast and narrowcast BW

56. Migration Strategy

57. DOCSIS 3.0 Migration Steps - Phased Approach for Improved Time-to-Market
Downstream Bonding
IPV6
Upstream Bonding
Multicast QoS
AES
IPDR

58. Initial Migration Goal
Deliver very high speed data service
Deliver 100+ Mbps DS
Deliver 50+ Mbps US
Reduction of node split cost
Multiple DSs per node
M-CMTS or I-CMTS load balancing
Multiple USs per node
Leverage existing ports and deploy 2.0 USs
BW flexibility & reduction of CMTS port cost
Break DS/US dependence i.e. independent scalability of US and DS
Reduce cost of DS ports by more than 1/10
Reduce CMTS port/subscriber cost by 30-50%

59. Migration Strategy Target CMTS upgrades in high priority markets
FiOS & U-Verse competitive markets
High growth & demographics
Markets with capacity issues
Your node 
Add more DS QAMs per service group and load balancing
Via I-CMTS and M-CMTS
Current 1x4 mac domain leaves US stranded
Increase capacity to existing 1.x/2.0 modem

60. Migration Strategy (cont)
Deliver targeted bonded DS channels to DOCSIS 3.0 CMs
Video and data convergence
Video and DOCSIS service group alignment
DSG & Tru2way will leverage DOCSIS DS BW
Share & leverage existing assets
UEQAMs for VoD, SDV and DOCSIS
UERM to enable QAM sharing

61. Cisco VDOC

62. What is VDOC?
Solution for the delivery of managed IPTV services over a DOCSIS network
Broadcast TV and VoD services
TV, PC, and other devices in the home
Provide user experience subscribers expect from their cable operator 62

63. IPTV – Even better on cable
Fat Pipes – DOCSIS 3.0
VBR video
IP/IP signaling/bearer channel as opposed to IP/MPEG
One Network (voice, video, data) to deliver them all
Delivery to alternate CPE outlets – PCs, Wifi PDAs (iPhone)
“Off-net” possibilities 63

64. Channel Bonding creates efficiency gains Big Channel “Packing Advantage”
No more room for HD
2 additional HD streams HD HD
Unbonded channels create inefficient boundaries
Bonding drives efficient “Packing”
Benefit varies
MPEG2/4 HD/SD mix
Bonding group size
10 SD + 5 HD streams SD HD
10 SD + 5 HD streams HD HD SD HD SD SD
Channel capacity HD SD SD HD SD SD HD HD SD SD HD SD SD SD SD HD SD SD SD SD SD SD 1 2 3 4 1 2 3 4
4 separate QAM channels
4-channel bonding group 64

65. Efficiency Gains from VBR Video
Support 40 – 60% more streams with VBR video
Law of large number works in favor of VBR statmux in fat pipe
As opposed to constant bitrate (CBR), VBR files vary the amount of output data per time segment. VBR allows a higher bitrate (and therefore more storage space) to be allocated to the more complex segments of media files while less space is allocated to less complex segments. The average of these rates can be calculated to produce an average bitrate for the file. 65

66. DOCSIS 3.0 Channel Bonding Concepts
A CM is unaware of the concept of bonding groups; it is only aware of the set of downstreams it must tune to and the flows it must forward, as instructed by the CMTS
A CM can receive traffic from multiple BGs simultaneously
Bonding groups may have different aggregate BW based on services supported, ie 1 BG = HSD and another BG = IPTV
Different CMs in a Service Group can receive traffic from different bonding groups, ie different BGs based on subscription levels
CM may tune to a subset of the downstreams configured for a SG
Number of receive channels on CM does not need to equal number of RF channels allocated to DOCSIS service (HSD/VoIP/IPTV) 66

67. Bonding Group Selection
A CM can receive traffic from multiple BGs
Operator can steer flows to particular BGs by configuring Service Flow attributes for each BG
CMTS uses SF-attributes when selecting BG for a flow
Operator could choose to set aside a BG for Cable IPTV and a separate BG for HSD/VoIP 67

68. DOCSIS 3.0 Channel Bonding Separate DS bonding groups for HSD/Voice and IPTV
CMTS
Integrated
or Modular
Service Group 1
HSD/VoIP
Video Headend
IPTV
IPTV System CM CM CM
STB / PC
STB / PC
STB / PC
Internet
IPTV
Service Group n
VoIP System
HSD/VoIP CM CM CM
STB / PC
STB / PC
STB / PC 68

69. RF Spanning Initial low-penetration IPTV deployments
CMTS
Integrated
or Modular
Service Group 1
HSD/VoIP
Video Headend
RF Spanning
IPTV System CM CM CM PC PC
STB / PC
IPTV
Internet
Service Group n
Similar to “wide and thin” in SDV deployments
VoIP System
HSD/VoIP CM CM CM PC PC
STB / PC 69

70. Cisco Architecture for D3.0 & M-CMTS

71. Cisco DOCSIS 3.0 DS Solution Deployed Worldwide Today
DOCSIS 3.0 Bronze functionality
Flexible M-CMTS Design
>2x DS capacity with incremental D3.0 module upgrade
40 to 184 DOCSIS DS ports
7Gbps CMTS Solution
DS channel bonding and narrowband currently supported on IOS 12.3(23)BC and 12.2(33)SCB
Compatible with all versions of the 5x20 including S,U, and H
US channel bonding supported in the Bighorn IOS release (FCS November 2009)
US channel bonding supported on the 5x20H, 3G60, 20x20
Supports >50,000 RGU’s per uBR10K
• Capacity for 24 DOCSIS Annex B and 18 Annex A downstream channels
• Integrated DOCSIS Media Access Control (MAC) processing
• Connectivity to QAMs using dedicated dual Gigabit Ethernet interfaces (SFP based)
• On-board packet-bonding engine that stripes IP packets across multiple DOCSIS downstream channels
The Cisco 1-Gbps wideband SPA performs all traditional and wideband DOCSIS processing of egress packets, including BPI+ encryption. The wideband MPEG packets are aggregated and encapsulated using the DOCSIS 3.0 packet-bonding technique (that is, User Datagram Protocol [UDP], IP, and Ethernet protocols [that is, packets are DOCSIS over wideband MPEG over UDP over IP over Ethernet]).
Expands the current 30-Mbps service portfolio options up to 240 Mbps
Provides a flexible and easy option to more than double the downstream capacity to the Cisco uBR10012.
Transcript:
I think most people are familiar with the SPA technology, this is what we call our SPA interface processor, the SIP card, code named Saratoga, this is a shipping SIP card. It's the card with which we've got the DOCSIS 3.0 Bronze functionality. It allows you -- like we were talking about flexible, modular CMTS design. It takes the uBR10K from 40 downstreams capacity today to 88, because you can have two of your downstream SPAs installed at 24 each, so 48 plus 40's 88. We support advanced 2.0 and 3.0 load balancing with the SPA technology and the SPA downstreams, full Layer 3 feature set packet cable and so on and so forth. The most interesting and operationally impacting feature of the SPA is that, like we briefly mentioned, the 5x20s and the RF cabling and wiring remains exactly the same on the 10K. The SPAs are add-on on the WAN side of the box. There's a GigE connectivity between the SPAs and your Edge QAM devices and it has very little operational impact on your existing services. And we support, as we said, advanced load balancing of 2.0 modems between your 5x20 downstreams and the add-on SPA downstreams at the same time that you might be running bonding services on the SPA downstream channels.
Author’s Original Notes:
Why Narrowband?
Continue to utilize the large installed base of legacy modems
Ability to increase the number of narrowband downstream channels in existing CMTSs
Simultaneous support for legacy modems at higher speeds
Support for 3-channel bonding
Ability to bond all channels of multi-channel CM
Additional downstream capacity
100 Mbps downstream data rates on a 3-channel modem
Provides support for mixed mode of operation for new QAMs

72. Cisco DOCSIS 3.0 DS Solution
Narrowband enables legacy DOCSIS [1.x/2.0] modems to use external QAMs for operation
Load Balancing and DCC techniques 1 – 4 are fully supported on SPA EQAM DS channels.
determine CM is an eMTA & initiate DCC to HA DS
Uses M-CMTS compliant Edge-QAM (EQAM) devices
Uses M-CMTS compliant DTI timing source for DS channels
Full Layer 3 IP routing feature set
Advanced QoS, VoIP, PCMM and MPLS VPN support for bonded services
The downstream Wideband channel is created by taking DOCSIS frames, putting them into MPEG-TS packets, and placing those MPEG-TS packets onto QAM carriers. However, instead of placing those MPEG-TS packets “horizontally” in time along a single QAM carrier as is done in traditional DOCSIS, the Wideband protocol places those MPEG-TS packets “vertically” across the QAM carriers assigned to a Wideband channel. A DOCSIS frame is literally tipped on its side and striped our across a group of QAM channels.

73. Cisco uBR10012 DOCSIS 3.0 Solution Reference Architecture

74. DOCSIS 3.0 Potential Option
5 DS freqs
3 US freqs
2x5 domain
Remote
DSs
Local
DSs
3.2 MHz
6.4 MHz
This diagram graphically shows fiber nodes vs frequency allocation with pros and cons of this scenario. This allows load balance of legacy modems between 2 DS frequencies and bonding on 4 DS frequencies. This requires 5 DS frequencies on the plant and 3 US frequencies.
We can use dcc tech 4 to load balance Basic subs across the local DS and one e-qam primary. If modems have a DS frequency in their config file, use “load balance exclude static strict” so only dynamic load balance takes place.
Pros
Four bonding freqs / e-qam connector
Only 1 e-qam connector per 8 nodes
Basic = 2 DS/2 nodes with DCC support
US load balance of 2.0 CMs
One US connector shared across 2 nodes for diminishing D1.x CMs
Cons
Requires M-CMTS architecture
Requires five DS & three US freqs
Must push 3.0 CMs to remote DS
Bonding group must be same IP bundle

75. DOCSIS 3.0 Option 1 Wiring Diagram
This diagram displays how DSs are potentially combined with e-qam modulator DSs and then split to feed multiple service groups or fiber nodes. Downstream four is not used and its associated USs are used in the four mac domains. Upstream connectors use internal frequency stacking on even connectors 0-14 and external stacking on connectors

76. Cisco DOCSIS M-CMTS

77. DOCSIS 3.0 Solution for the uBR7200VXR Series UBR-MC8x8U---Extending UBR7200 Series to DOCSIS3.0
Full DOCSIS 3.0 compliance
DS bonding/US bonding
Legacy DOCSIS 1.x and 2.0 modem support
Multicast, IPv6 and other DOCSIS 3.0 specs
S-CDMA and logical channels
AES encryption
Same form-factor as current UBR-MC28U line card, upgrade is simple LC swap
Operates in 8 DS/8 US mode on UBR7225VXR and UBR7246VXR, 4x DS density of the existing MC28U line card
Requires UBR7200-NPE-G2 77

78. DOCSIS 3.0 evolution with the UBR10k
MC520H with D3.0 SPA
88 DS solution with DS bonding
MC520H with 6 D3.0 SPA, PRE4 and 10G
184 DS solution enables 5+ DS per FN
US Bonding on the MC520H
Enables higher US rate service offerings
MC2020
Full D3.0 capability and line rate US bonding
Easy upgrade from 520H; interoperable with the D3.0 SPA
MC3G60
Enables 8+ channel DS bonding at scale
Scales US by 3x

79. US Channel Bonding on MC520H
DOCSIS 3.0 2, 3, and 4 channel US bonding supported
100 Mbps throughput on US bonded flows per line card
DOCSIS Line rate on D2.0/Non-bonded CM
BPI+ and PHS support for 3.0 and 2.0 flows
Dynamic BW sharing between 2.0 and 3.0 flows
Feature supports provisioning 3.0 CM in bonded or non-bonded configuration
Different US rates supported in Bonding Group
For example: 16QAM/3.2Mhz + 64QAM/6.4Mhz

80. Cisco uBR10K MC2020 Linecard Full DOCSIS 3.0 support DSCB USCB IPv6
MCast
AES
Upgrade for MC520 LCs
Same RF Cabling
Very low operational impact
Greater than 7x DS capacity in same 10K footprint
Grow from 40 DSs to 304 DSs with MC2020 and six D3.0 SPAs
>10Gbps CMTS solution
Full HA support 80

81. MC2020 Features Full DOCSIS 3.0 compliance
DS bonding/US bonding
Legacy DOCSIS 1.x and 2.0 modem support
Multicast, IPv6 and other DOCSIS 3.0 specs
S-CDMA and logical channels
AES encryption
Line rate performance on US and DS on all channels (Annex A/B)
MC2020 as Protect for MC520 and MC2020
Full Feature parity with MC520
PRE2/PRE4 support
Interoperable with the DOCSIS 3.0 DS SPA
SW licensing
0x20V, 5x20V, and 20x20v SKUs
5 DS, 15 DS, and 20 DS upgrade licenses will be made available

82. MC2020 with MC520H in the same UBR10K chassis
MC2020 in 2 slots configured as “Working”
1 MC2020 configured as “Protect”
MC520H occupy other RF slots (“Working”)
MC2020 acts as Protect for BOTH MC520H/MC2020
SPA slots can be occupied by 6 D3.0 DS SPA
Slots Filled
DS Spigots
DS Channels
MC520H 5
25 (5 * 5) 25
MC2020 2
10 (2 * 5) 40
D3.0 SPA
6 (SPA Slots)
6 GigE
144
MC2020 as Protect
For 520H and 2020
Total DS channels in this configuration
= 209

83. Cisco uBR10K MC3G60 Linecard
Greater than 12x DS capacity in same uBR10K installed chassis
576 DS (504 DS with HA)
~20Gbps DOCSIS connectivity
10Gbps backhaul
3x US capacity
480 US (420 US in HA)
Up to 12:1 freq stacking on US ports
Scalable and efficient uBR10K and RFGW-10 matching
Full HA on 10K and RFGW-10
MC3G60
MC3G60
MC3G60
MC3G60
MC3G60
MC3G60
MC3G60
MC3G60 US GE DS
RFGW-10 83

84. 3G60 Highlights Full DOCSIS 3.0 compliance
Line rate DS bonding/US bonding
Legacy DOCSIS 1.x and 2.0 modem support
Multicast, IPv6 and other DOCSIS 3.0 specs
S-CDMA and logical channels
AES encryption
DEPI M-CMTS
15 Mac Domains per LC
72 DS channels and 60 US channels
N+1 LC redundancy
Flexible US and DS ratios (4/8/16/24 channel DS bonding)
SW licensing options

85. Bandwidth Growth / Capacity Transition Points 10K Migration
20x20
Spumoni
Saratoga
uBR10K scales well ahead of maximum bandwidth demand
3G60 supports high-capacity V-DOC in 1 chassis through 2015

86. Summary

87. New Technology Cornerstones
DOCSIS channel bonding for higher capacity
Enable faster HSD service
MxN mac domains now
Enable video over IP solutions
M-CMTS
Lower cost downstream PHY
De-couple DS and US ports
I-CMTS
Allows higher capacity in same box
Same wiring
New technologies are being pursued to address the DS bottleneck conundrum.
DOCSIS 3.0 uses a channel bonding technique to achieve higher capacity links, enable faster high speed data (HSD) service, and provide M x N MAC domains to enable video over IP solutions.
The idea of multiple grants per request or outstanding requests with DOCSIS 3.0 is also a good idea for US speed, but still doesn't fix the CPU issue on the CM.
The modular CMTS (M-CMTS) architecture is promoted to achieve better DOCSIS economics, lower cost DS PHY, and de-couple DS and US ports. One day we may see fiber optic nodes with DOCSIS physical layer chips embedded so we can use ingress cancellation at the node, digital links from the node back to the headend without the need to amplify, and no more laser clipping. Of course, this means all traffic needs to be DOCSIS-based on the US!
The integrated-CMTS idea is a way to use “off-the-shelf” external QAM boxes and the existing CMTS just for US connectivity. This allows return on invest and no “fork-lift” upgrades of the CMST chassis and/or limited cabling changes.

88. DOCSIS 3.0/M-CMTS Concluding Remarks
Promises ten times BW at fraction of cost
Introduce new HSD service of 50 to 75 Mbps
Backward compatible with existing DOCSIS standards
Allows migration of existing customers to higher tier and DOCSIS 3.0 capability
Allows more BW for legacy DOCSIS 2.0 CM
Allows for a phased deployment
IPV6, US bonding, and other features will follow
DOCSIS 3.0 enables a media-rich user experience with:
Faster speeds
More devices
Service convergence
DOCSIS 3.0 provides the tools required to build a scalable, efficient, cost-effective access network
Cisco DOCSIS 3.0 solutions provide industry-leading scalability, density and cost performance