DOCSIS 3.1: PLANS AND STRATEGIES December 18, 2013
SCTE LIVE LEARNING Monthly Professional Development service Generally Hot Topics or Topics of high interest to the industry Vendor Agnostic No product promotion Free to SCTE members Live sessions are recorded Members-only benefit
TODAY S SESSION Approximately 50 minutes discussion 10 minute Q&A at the end, however.. Ask questions anytime throughout the session Asking questions adds value and enhances learning opportunity for you and others
HOW TO ASK A QUESTION Select the Q&A tab Type in your question and submit
NOW LET S GET STARTED DOCSIS 3.1: PLANS AND STRATEGIES December 18, 2013
Patricio Latini Cable Consultant and (former) Chief Regional Technologist Latin America ARRIS Managed the technological strategy for ARRIS in the Latin American region Has written more than 20 technical articles and has spoken at the most important conventions of the industry More than 15 years of experience in the cable industry leading the deployment of IP voice and data services over CATV networks
AGENDA Introduction BW Trends in Next-Generation Networks HFC Network Evolution to DOCSIS 3.1 Summary
INTRODUCTION We are entering an interesting era when MSOs will begin to: Experience the exponential growth of DOCSIS HSD Offer 1+ Gbps service to subscribers Make major changes to their networks Monitor subscriber quality of experience to guide future paths Develop plans for the long-term evolution of their fiber networks
AGENDA Introduction Bandwidth Trends in Next- Generation Networks HFC Network Evolution to DOCSIS 3.1 Summary
DOCSIS (HSD): FASTEST-GROWING & HIGHEST PROFIT MARGIN IN HFC SPECTRUM It must be managed with special attention The Nielson Curve for Cable (Max. DS Capacity/sub) Max DS Permitted for Modems (bps) 100G 10G 1G 100M 10M 100 10 The past 25-years show a constant bandwidth increase of ~1.5x every year. 200 Mbps 12 Mbps 1M 1 Mbps 5 Mbps 100k 256 kbps 512 kbps 56 kbps 28 kbps 128 kbps 10k 9.6 kbps 33 kbps 14.4 kbps 1.2 kbps 1k 2.4 kbps 300 bps 4.3 Gbps DS Limit (750 MHz) The Era of Dial-Up Modems 50 Mbps The Era of Cable Modems The Era of Wideband Cable Modems ~73 Gbps? 50% CAGR for 30+ years Assumes average BW also grows with A 50% CAGR 1 1982 1986 1990 1994 1998 2002 2006 Year 2010 2014 2018 2022 2026 2030 2023
PER-SUBSCRIBER AVERAGE BANDWIDTH TRENDS Avg Busy-Hour DS BW per Subscriber (kbps) DS:US BW Ratio Avg Downstream BW per Subscriber 16000 14000 12000 10000 8000 6000 4000 2000 0 2005 2010 2015 2020 2025 Year DS:US BW Ratio 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2005 2010 2015 2020 2025 Year Assume a 50% CAGR MSO A DS BW/Sub (CAGR=69%) MSO B DS BW/Sub (CAGR=43%) MSO C DS BW/Sub (CAGR=57%) MSO A DS:US BW Ratio MSO B DS:US BW Ratio Avg Busy-Hour US BW per Subscriber (kbps) DS:US Channel Ratio 400 350 300 250 200 150 100 50 80.0 60.0 40.0 20.0 Avg Upstream BW per Subscriber 0 2005 2010 2015 2020 2025 Year DS:US Channel Ratio MSO C DS:US BW Ratio 0.0 2005 2010 2015 2020 2025 Year Assume a 20% CAGR MSO A US BW/Sub (CAGR=42%) MSO B US BW/Sub (CAGR=19%) MSO C US BW/Sub (CAGR=11%) MSO A DS:US Ch Ratio MSO B DS:US Ch Ratio MSO C DS:US Ch Ratio
DRIVERS OF BANDWIDTH GROWTH Growth in existing service types Web-Browsing: Average web pages are currently 1.3 Mbytes & are experiencing 35% CAGR IP Video: H.264 1080p HD video feeds currently consume ~8 Mbps 4K UHD video feeds will consume 4x the BW of 1080p without improved compression ~32 Mbps 4K UHD video feeds will consume 2x the BW of 1080p with HEVC compression ~16 Mbps 8K UHD video feeds will consume 16x the BW of 1080p without improved compression ~128 Mbps 8K UHD video feeds will consume 8x the BW of 1080p with HEVC compression ~64 Mbps Arrival of entirely new service types Ex: 3DTV, holographic TV, etc.
CAN 50% DS GROWTH CONTINUE? There are several philosophical views on this Home Concurrency Streams/ Active Home Avg BW/Stream Highest Avg BW/Home View 1 0.5 1.5 7 Mbps (H.264 HD) 5.25 Mbps (fills human senses on TV) View 2 0.5 1.5 56 Mbps (H.265 UHD) 42 Mbps (fills human senses on wall) View 3 &4 Not Applicable Not Applicable Not Applicable No Limit (machine to machine) Assume Average BW/Home = 300 kbps today Assume 50% Annual HSD growth today HSD growth may go until 2019 (View 1) or 2024 (View 2) or forever (View 3) or could taper (View 4) Per-Subscriber Bandwidth (Mbps) 140 120 100 80 60 40 20 0 2010 2015 2020 2025 Year View #1 View #2 View #3 View #4
DOCSIS (HSD): FASTEST-GROWING & HIGHEST PROFIT MARGIN IN HFC SPECTRUM It must be managed with special attention DS BW as a function of time (w/ ~50% Annual Growth Rate) DS BW for Modems (bps) 100G 10G 1G 100M 10M 1M 100k 10k 1k 100 10 D3.1 DS Limit = 10.8 Gbps (1200 MHz) D3.0 DS Limit = 4.9 Gbps (750 MHz) 300 bps Nielson Law s Tmax The Era of Dial-Up Modems The Era of Cable Modems OFDM BW for 64 Subs/SG The Era of Wideband Cable Modems ~100 kbps in 2010 Avg BW/SG The Era of 3.1 Modems How can we ensure an MSO has an easy path to here & beyond? Tavg/sub 512 Subs/SG 256 Subs/SG 128 Subs/SG 64 Subs/SG 64 DOCSIS 3.0 Chans. 32 DOCSIS 3.0 Chans. 16 DOCSIS 3.0 Chans. ~332 Mbps? 4 DOCSIS 3.0 Chans. 1 DOCSIS 3.0 Chan. 1 1982 1986 1990 1994 1998 2002 2006 Year 2010 2014 2018 Today 2022 2026 2030
UPSTREAM BW NUMBERS WILL ALSO CONTINUE TO GROW AT A SLOWER GROWTH RATE US BW as a function of time (w/ ~20% Annual Growth Rate) How can we ensure an MSO has an easy path to here & beyond? 100G Avg BW/SG 10G 1G 100M 44 DOCSIS 3.1 Chans. (5-293 MHz) 12 DOCSIS 3.0 Chans (5-85 MHz) 6 DOCSIS 3.0 Chans. (5-42 MHz) 512 Subs/SG 256 Subs/SG 128 Subs/SG 64 Subs/SG US BW for Modems (bps) 10M 1M 100k 10k 1k 100 10 1 1 DOCSIS 3.0 Chan. The Era of Dial-Up Modems 1982 1986 1990 1994 1998 Tmax? 10 Mbps in 2010 The Era of Cable Modems 2002 2006 Year OFDM BW for 64 Subs/SG ~30 kbps in 2010The Era The Era of of 3.1 Wideband Modems Cable Modems 50% 30% 10% 2010 2014 Today 2018 Tavg/sub 2022 2026 2030 ~1.15 Mbps? One 192 MHz OFDM block may be adequate yet two may be desired
AGENDA Introduction BW Trends in Next-Generation Networks HFC Network Evolution to DOCSIS 3.1 Summary
A QUICK REVIEW: THE HISTORY OF DOCSIS DOCSIS 1.0 March 1997 Data Over Cable System Interface Specification (DOCSIS) begins Defined support for high-speed data over HFC DOCSIS 1.1 April 1999 Added state of the art QoS techniques for priority services (e.g. VoIP) DOCSIS 2.0 December 2001 Increased upstream modulation format for more b/s/hz Added new PHY for the upstream SCDMA DOCSIS 3.0 August 2006 Increased BW capacity with Channel Bonding Added IPv6 & Multicast QoS DOCSIS is ready for another update. What features are most important to MSOs now?
DOCSIS 3.1: HIGHER SPECTRAL EFFICIENCIES DOCSIS 3.1 goals stated by CableLabs at SCTE Cable Tec Expo 2012 ARRIS is on all committees. DOCSIS 3.1 Spec completion planned for this year DOCSIS 3.1 aims to augment DOCSIS with PHY & MAC technology that: Increases Spectral Efficiency (bps per Hz) of the plant by 20-50% Permits 10+ Gbps of downstream BW capacity on the HFC plant Permits 1+ Gbps of upstream BW capacity on the HFC plant Reduces the cost per Mbps (relative to DOCSIS 3.0) Adapts to different amounts of spectrum and plant conditions Permits an easy migration strategy from the current architecture Operates on existing HFC networks and active components Supports energy efficiency features such as low power standby and/or sleep modes
DOCSIS 3.1 TO THE RESCUE! Capitalizes on new LDPC FEC & OFDM PHY technologies Permits higher modulation orders (1024 QAM, 4096 QAM, etc.) Should permit spectral efficiency increases to be much closer to the Shannon-Hartley Limit S/N(dB) Eliminates 6 MHz & 8 MHz channelizations (NA & Europe can unify) Must be backwards compatible with previous DOCSIS standards, supporting at least 24 legacy DSs and 8 legacy USs per service group Must support high-split US operation up to at least 200 MHz Must support DS operation up to at least 1.2 GHz Will use bit-loading to adjust to the HFC plant S/N Max. Spectral EfficiencyLimit (bps/hz) 36 3981 11.96 33 1995 10.96 30 1000 9.97 27 501 8.97 24 251 7.98
DOCSIS 3.1 TECHNICAL DIRECTIONS Expanded DS & US Spectrum High-Split Architecture NEW PHY Technology OFDM DS OFDMA US NEW FEC Technology LDPC
DOCSIS 3.1 TECHNICAL DIRECTIONS Multiple DS modulation profiles for efficient performance Broadcast strategy Physical Layer channel In-band channel: preamble, timestamp & sync info, boot-up info, etc. Worst Case Average Case Best Case CW pointer channel In-band channel: pointers to FEC CWs boundaries
WHY HIGH-SPLIT ARCHITECTURE? The Good Offers the highest US system capacity Less signal attenuation from cable loss Single transition band is required The transition band is narrow because it occurs at low frequency Does not place a limit on the growth of the DS spectrum Offers backward compatibility for legacy devices with 5-85MHz support Offers the most cost-effective solution The Ugly It affects Out-of-Band (OOB) STB signaling (~70 130 MHz) DS spectrum layout and channel assignments are affected Requires changing most actives (Top-Split requires changing all actives)
POLL QUESTION What percentage of your network already supports upstream bandwidth above 42 MHz? 0 % 25 % 50 % All my network is in either 65 MHz or 85 MHz Respond to the survey in the window on the right and click Submit
A TAXONOMY OF POSSIBLE, FUTURE MODULATION TECHNIQUES Modulation Techniques Single-Carrier (Legacy DOCSIS) Multi-Carrier (DOCSIS 3.1) QAM OFDM
WHAT IS OFDM? OFDM Orthogonal Frequency Division Multiplexing Multi-Carrier Technology Composed of Subcarriers FFT-based Implementation Widely adopted in the OFDM Single-Carrier OFDM Subcarriers
MULTI VS. SINGLE-CARRIER SYSTEMS Single Carrier (ex: J.83 of DOCSIS 3.0 & EQAMs): symbols sent sequentially at fast symbol rate Multi-Carrier (ex: OFDM of DOCSIS 3.1 ): symbols sent in parallel at slow symbol rate Ingress Noise Frequency Frequency Impulse Noise S 4 S 3 S 2 S 0 S 1 S 2 S 3 S 4 S 1 S 0 Single Carrier System 1. QAM symbols sent sequentially 2. Short symbol periods (susceptible to impulse noise) ~ <1 µs 3. Each QAM symbol modulates the whole spectrum (susceptible to ingress noise) Multi-Carrier System 1. QAM symbols sent concurrently 2. Long symbol periods (robust against impulse noise) ~ 20-40 µs 3. Each QAM symbol modulates a single subcarrier (robust against ingress noise)
POWER OF INDIVIDUAL OFDM SUBCARRIERS Individual subcarriers are managed individually Each subcarrier can be turned on/off Each subcarrier can have its own modulation order (variable bit-loading) OFDM Channel with nulled subcarriers to accommodate legacy channels
SUBCARRIERS CAN BE TURNED ON/OFF! Mitigates the effect of narrowband interference 5 x 10-7 7 x 10-8 4 Few active OFDM subcarriers 3 2 1 0-1 -2 6 5 4 3 2 OFDM channel with no gap -3 1-4 -5-3 -2-1 0 1 2 3 0-3 -2-1 0 1 2 3 (a) (b) 5 x 10-7 7 x 10-8 4 Many active & few nulled OFDM subcarriers 3 2 1 0-1 -2-3 6 5 4 3 2 OFDM channel with a gap -4 1-5 -20-15 -10-5 0 5 10 15 20 0-20 -15-10 -5 0 5 10 15 20 (c) (d)
BENEFITS: MANAGING INDIVIDUAL SUBCARRIERS Individual subcarriers experience different SNR values OFDM s variable bit-loading permits the selection of appropriate modulation order for each subcarrier Narrowband Interferer SNR Frequency Variable bit-loading: various subcarriers can have different QAM modulation orders 1024 QAM 256 QAM 1024 QAM Mute 1024 QAM
BENEFITS: MANAGING INDIVIDUAL SUBCARRIERS Deployed DS channel characteristics vary at roll-offs & above 1 GHz Each modem may experience different spectral responses OFDM s variable bit-loading permits the selection of appropriate modulation order for each subcarrier 0 Insertion Loss (db) -10-20 -30-40 -50-60 0 500 1000 1500 2000 2500 3000 Frequency (MHz)
LDPC BASICS Low Density Parity Check Much more efficient than traditional Reed-Solomon codes (RS) FEC codes used in DOCSIS Invented in 1960s by Gallager, but not used until recently because of implementation complexity LDPC codes are in many standards including DVB-C2, WiMAX, MoCA, G.hn *Figure obtained from Algorithms and Architectures for Efficient Low Density Parity Check (LDPC) Decoder Hardware, Ph.D. dissertation by TINOOSH MOHSENIN, UC Davis, 2010.
LDPC PERFORMANCE IN DVB-C2 Getting very close to Shannon limit 256-QAM 64-QAM 128-QAM 32-QAM 16-QAM
EXAMPLE: LDPC GAIN COMPARED TO RS FEC 0.92 Code rate 0.9 0.88 0.86 0.84 0.82 LDPC offers an average SNR gain of 5.5 db in the QAM256 case* 0.8 0.78 0.76 QAM256, LDPC FEC (DVB-C2) QAM64, LDPC FEC (DVB-C2) QAM256, RS FEC QAM64, RS FEC Simulation for BER = 10-10 0.74 16 18 20 22 24 26 28 30 SNR (db) *Assuming large US traffic that enables the use of large FEC code blocks
FREQUENCY & TIME INTERLEAVING Frequency interleaving is needed to mitigate the effect of narrowband interference (ingress, LTE, etc.) Distributes errors occurring to consecutive subcarriers over multiple FEC CWs May not need to null subcarriers Performed at the subcarrier level Time interleaving helps in mitigating the effect of impulse noise Distributes the errors covering an entire OFDM symbol over multiple OFDM symbols Performed at the subcarrier level Frequency Interleaver Time Interleaver
FREQUENCY & TIME INTERLEAVING Multiple types of interleavers can be used Block interleaver (will be used for Frequency) Writes along columns and reads along rows Convolutional Interleaver (will be used for Time) Requires ½ memory size & ½ latency compared to block interleaver Subcarriers experience different delays at the interleaver and de-interleaver, nevertheless, all subcarriers experience identical total latency Interleaver De-Interleaver
INTERLEAVING No Interleaving Time Interleaving Frequency Interleaving Time and Frequency Frequency Frequency Frequency Frequency Time No Protection against Burst or Narrowband Noise Time Protection Against Burst Noise Time Protection Against Narrowband Interference Time Protection against Burst or Narrowband Noise
POLL QUESTION What is the average downstream SNR of the nodes in your network? 34 db 36 db 38 db Over 39 db Respond to the survey in the window on the right and click Submit
DOWNSTREAM EXAMPLE: LIMITS OF CURRENT RF DATA TECHNOLOGY Millions of Cable Modems DOCSIS 3.1 Possible Modulation Formats Mean = 36.6 db Improvement vs. Today Histogram of Cable Modem Downstream Downstream Signal to Noise Ratio in db DOCSIS / ITU-T J.83 Annex A: 256 QAM 1024 QAM 8/9 Coded Req. Min 33 db DOCSIS / ITU-T J.83 Annex B: 256 QAM 2048 QAM 8/9 Coded Req. Min 36 db 4096 QAM 8/9 Coded Req. Min 39 db 8192 QAM 8/9 Coded Req. Min 42 db 16384 QAM 8/9 Coded Req. Min 45 db ~30.4% ~43.6% ~56.7% ~63.2% ~71.2%
UPSTREAM EXAMPLE: LIMITS OF CURRENT RF DATA TECHNOLOGY Mean = 32.9 db SD = 2.15 db 24 26 28 30 32 34 36 38 Upstream SNR (db) DOCSIS 64QAM ATDMA DOCSIS 3.1 Possible Modulation Formats 512 QAM 0.85 Coded Req. Min 27 db 1024 QAM 0.75 Coded Req. Min 29 db 1024 QAM 0.85 Coded Req. Min 31 db 1024 QAM 0.85 Coded Req. Min 33 db 1024 QAM 0.89 Coded Req. Min 35 db 2048 QAM 0.85 Coded Req. Min 38 db Improvement vs. Today ~46.0% ~48.4% ~67.8% ~67.8% ~75.9% ~93.5%
THE BANDWIDTH CAPACITY POTENTIAL OF DOCSIS (WITH NEW TECHNOLOGIES) Now Phase 1 Phase 2 Phase 3 DS Range (MHz) 54-1002 108-1002 300-1152 500-1700 DS QAM Level 256 256 1024 1024 # DS Channels 8 24 142 200 DS Capacity (bps) 300M 1G 7G 10G US Range(MHz) 5-42 5-85 5-230 5-400 US QAM Level 64 64 256 1024 # US Channels 4 12 33 60 US Capacity (bps) 100M 300M 1G 2.5G Note: TBD values are underlined Channels in quotes = Equivalent number of Single Carrier QAMs
SAMPLE HFC MIGRATION PLAN IN THE DOCSIS 3.1 ERA Phase 3 Up Down Legacy Video EQAM (Digital Video) CMTS & EQAM DOCSIS 1.0-3.0 (HSD, VoIP, & IP Video) CCAP DOCSIS 1.0-3.0 DOCSIS 3.1 in US DOCSIS 3.1 in DS (HSD @ PON Speed, Video over IP, Ultra HD, & un-discovered apps) Time
AGENDA Introduction BW Trends in Next-Generation Networks HFC Network Evolution to DOCSIS 3.1 Summary
SUMMARY Traffic growth is driven by demand and competition The DOCSIS 3.1 spec will greatly increase the performance of HFC networks via BW extension, OFDM PHY, & LDPC FEC 10+ Gbps DS & 2+ Gbps US will permit DOCSIS to satisfy subscriber BW needs well into the 2020 decade DOCSIS scales very well: offers just-in-time investment steps RFoG and/or Digital Optics can be used with DOCSIS 3.1 for increased offered capacities DPoE Specification leverages existing OSS in PON like services The cable industry is in a excellent position and off to a great journey! MM14
ACKNOWLEDGEMENTS Thanks to Ayham Al-Banna and Tom Cloonan from ARRIS for contributing with material for this Presentation
HOW TO ASK A QUESTION Select the Q&A tab Type in your question and submit
THANK YOU TO OUR SPEAKER Patricio Latini Consultant platini@gmail.com
REMINDER This session has been recorded Will be available on SCTE s Member s Only Site within 2-3 days To access previously recorded sessions login to: www.scte.org with your member ID#, then scroll to the bottom of the page and select SCTE Live Learning Archives for a menu of previously recorded Live Learning sessions
LIVELEARNING ARCHIVES Free for SCTE Members www.scte.org Under Resources/LiveLearning Archives Topics include: Advanced Advertising Broadband Premises Business Services DOCSIS Emergency Alert System (EAS) Energy Ethernet Fiber Transport HFC Systems Home Networking IP OCAP PacketCable Routing Service Management Standards Video VoIP Wireless Technology
NEXT MONTH Register for the next SCTE Live Learning webinar TBD January 15, 2014 2:00 p.m. Eastern www.scte.org Under Professional Development/Live Learning Available today at the conclusion of this presentation
LIVELEARNING REGISTRATION Register for the SCTE LiveLearning Series www.scte.org Under Professional Development/ LiveLearning Webinars Available today at the conclusion of this presentation Third Wednesday of the month at 2 PM Eastern