RFoG. Beginnings, today, and what the future holds. SCTE Piedmont Chapter, July 2015 Douglas Pieri, Staff Solutions Engineer.
Download ReportTranscript RFoG. Beginnings, today, and what the future holds. SCTE Piedmont Chapter, July 2015 Douglas Pieri, Staff Solutions Engineer.
RFoG. Beginnings, today, and what the future holds. SCTE Piedmont Chapter, July 2015 Douglas Pieri, Staff Solutions Engineer RFoG History • RFoG is the result of SCTE’s Interface Practices Subcommittee Working Group 5 That defined the specifications and operation of the subscriber ONU device. • “Defines a fiber-to-the-home system optimized for compatibility with HFC plant, using the same end equipment at both the home and the HE or Hub.” Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Traditional HFC compared to RFoG 1 GHz HFC System at 256 HHP Per Physical Node Coax Segment ~64 HHP Per Segment 54 MHz -1 GHz 5 MHz - 42 MHz 1 GHz RFoG System at 32 HHP Per Optical Segment 54 MHz -1 GHz 1 – 32 HHP 1550 nm WDM Mux 5 MHz - 42 MHz 1610 nm Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Uses Existing Coax Wiring & CPE “just like today” 1550 nm // 1610 nm Splitter Still HFC, but… Fiber Extends All the Way to the Home WDM Mux DOCSIS Set-Top DOCSIS CM/EMTA 3 Traditional PON compared to RFoG 1 GHz RFoG System at 32 HHP Per Optical Segment 54 MHz -1 GHz 20km 1550 nm WDM Mux 1610 nm 5 MHz - 42 MHz +20dBm Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 1 – 32 HHP 1550 nm // Splitter 1610 nm WDM Mux -17.0 +18.5dBm -3.5dBm 4 RFoG with Repeater Architecture 1 – 32 HHP 1550 nm // Splitter 1 – 32 HHP 1610 nm 1550 nm // 54 MHz -1 GHz RFoG Repeater WDM Mux W D M Splitter 1610 nm W D M WDM Mux WDM Mux 256HP Typical 5 MHz - 42 MHz 1 – 32 HHP Closely Resembles Traditional Node 1550 nm // Splitter 1 – 32 HHP 1610 nm 1550 nm // Splitter Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 1610 nm WDM Mux WDM Mux 5 RFoG Architectures Copyright 2015 ARRIS Enterprises, Inc. All rights reserved. ARRIS–Confidential & Proprietary 6 Architectures • Several Architectures have been developed. Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Architectures Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Architectures • Pros • Cons • Matches traditional PON architectures. • Expensive in very low densities. • Concurrent operation with PON technologies. • Wastes optical power in low density. • Strands optical power in low penetrations. Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Architectures • Pros • Cons • Maintains split ratio of 1x32. • Expensive in very low densities. • Concurrent operation with PON technologies. • Wastes, expensive optical power in low density. • Leverages lower penetrations better. • Strands optical power in low penetrations. • Can cause long term cross connect confusion. Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Architectures • Pros • Cons • Maintains split ratio of 1x32. • Creates high fiber counts in the network. • Concurrent operation with PON technologies. • Requires field cabinets, vs Splice enclosures. • Leverages lower penetrations best Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. Architectures • Pros • Cons • Utilizes Optical power more effectively. • Does not maintain 1x32 split ratio. • Upstream can be more balanced in low densities. • Multiple failure points. • Difficult to troubleshoot network issues. • Resembles traditional HFC. • Creates very low fiber counts Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. RFoG Architecture Benefits Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. OBI Mitigation Comparison Copyright 2015 ARRIS Enterprises, Inc. All rights reserved. ARRIS–Confidential & Proprietary 14 Optical Beat Interference Clean Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. OBI • Optical Beat Interference (OBI) can occur when two or more RFoG ONUs (Optical Network Units) transmit simultaneously to the same receiver, and contain wavelengths which are sufficiently close to each other • OBI is a result of the heterodyning of the two (or more) closely spaced wavelengths present on the same detector. Heterodyning results in the downmixing of the optical frequencies of the two or more lasers into the RF domain, appearing as wideband noise. • The resultant noise has the ability to impact the signal integrity of upstream communication channels, most notably on the receiver where the OBI has occurred, but can also affect the upstream DOCSIS service group via the introduction of noise into the RF combining network presented to the CMTS blade 15 Ways to Control OBI • Manage upstream Wavelength from ONU – Requires l spacing > 0.5nm with precision l control, tunable lasers, or multiple SKU’s – Must maintain database of l assignments per user • Manage upstream bursts in time domain (Scheduler) – CMTS limits to one-ONU-at-a-time, prevents overlap – Limits efficient utilization of channel capacity – Efficiency constraints become increasingly serious for high capacity upstream configurations (more channels, more bandwidth, more utilization) – a serious concern! • Existing techniques have limitations that may become severe in the future! Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 16 Wavelength Management WL Drift (nm) Start-up Drift l management approach requires spacing >0.5nm to assure zero OBI 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0 100 200 300 400 Laser Startup Time (us) Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 500 17 Normal CMTS 3.0 Bonded Upstream Operation CMTS Ch 1 Ch 2 Ch 3 Ch 4 Upstream Fiber MAPs CM #1 (3.0) ONU 1 CM #2 (2.0) ONU 2 CM #3 (3.0) ONU 3 Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. l1 l2 l3 18 Burst Management CMTS Scheduler (Packet Alignment in CMTS MAPPER) CMTS Ch 1 Upstream Fiber MAPs Ch 2 Ch 3 Ch 4 CMTS Scheduler with Single ONU TX at a time Can Suffer From Inefficient Utilization Of The Channel BW CM #1 (3.0) ONU 1 CM #2 (2.0) ONU 2 CM #3 (3.0) ONU 3 Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. l1 l2 Becomes worse with increased Upstream Capacity, especially D3.1 l3 19 Burst Management DOCSIS 3.0 Channel Efficiencies Std CMTS Sched w/ OEO vs. CMTS Single TX OBI Sched – DOCSIS 3.0 Channel Efficiency 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% • OEO AgileMax CMTS OBI Sched 1 2 3 4 5 6 7 8 # of DOCSIS 3.0 Upstream Channels Note: ~500B average packet size with 70% small, 30% large packets; all 3.0 modems Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 20 Active – OBI Elimination … … Repeater HFC Headend HFC Segment Single Family RFoG Long Reach supports PON l s Requires Power ~8 Watts per 32 ports … Hub or Field Location Replaces passive splitter Long Reach MDU PON Fiber Splitter OBI completely eliminated with standard ONUs & no special scheduler! Full downstream & upstream throughput for DOCSIS 3.0 and 3.1 Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 21 April 2015 21 Active – OBI Elimination Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 22 EPON Inefficiency Copyright 2015 ARRIS Enterprises, Inc. All rights reserved. ARRIS–Confidential & Proprietary 23 EPON Scheduler Burst US Structure Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 24 10G EPON Upstream Analysis • Industry Literature shows EPON Upstream impacted by # of ONU/LLID and Grant Cycle Time – See several papers by Glen Kramer, Broadcom/Teknovus and Marek Hajduczenia, BrightHouse (formerly with ZTE) 1 ms 2 ms 4 ms 8 ms 32 85.00% 86.05% 86.57% 86.84% 64 82.91% 85.00% 86.05% 86.57% 128 78.72% 82.91% 85.00% 86.05% 1 ms 2 ms 4 ms 8 ms 32 8.47 Gbps 8.59 Gbps 8.65 Gbps 8.68 Gbps 128 7.78 Gbps 8.24 Gbps 8.48 Gbps 8.59 Gbps ONUxLLID ONUxLLID Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 25 EPON Upstream Cycle Time and ONU/LLID Impacts 10G EPON US (ONUs, LLIDs, Cycle Time) 1G EPON US (ONUs, LLIDs, Cycle Time) 10 1000 Cycle Time Capacity (Gbps) 8 7 8 6 4 5 2 4 1 3 0.75 2 0.5 1 0 900 Capacity (Mbps) 9 800 8 700 4 600 2 500 1 400 0.75 300 0.5 200 100 0 32 64 128 256 512 1024 32 64 Total Active ONU x LLID • 128 ONU, 4 Active LLID, 1msec Cycle – Only 50% Efficiency Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 128 256 512 Total ONUxLLIDs • 64 ONU, 4 Active LLID, 1msec Cycle – – Only 50% Efficient Similar to D3.1, 85MHz return 26 RFoG: High Capacity on One Fiber A single fiber infrastructure supporting RF/DOCSIS, 10G EPON, and 10G Ethernet. Technology EPON RF/DOCSIS Ethernet Total Aggregate Capacity in GB/s (DS/US) Near Term Future 10/10 40/40 (really 4x 10/10) 5.6/0.3 12.5/1.8 (D3.1) 10/10 (ENS) 20/20 (or more) 25.6/20.3 72.5/61.8 FWD FWD HP ONU RTN RTN LP RTN FWD Coax HP LP FWD RTN RTN A2D CPU D2A CPE And beyond that… Future CPEs willrequires allow RFDS Spectrum “re-use” removing Coax from the equation Today the Coax and US to occupyby different portions of the Spectrum Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. RFoG: High Capacity on One Fiber A single fiber infrastructure supporting RF/DOCSIS, 10G EPON, and 10G Ethernet. Technology EPON RF/DOCSIS Ethernet Total Aggregate Capacity in GB/s (DS/US) Near Term Future Closer 10/10 40/10 (really 4x 10/10) ‘Future’ 5.6/1.0 40/10 (D3.x) than PON 10/10 (ENS) 20/20 (or more) 25.6/21.0 100/50 FWD FWD HP ONU RTN RTN LP RTN FWD Coax HP LP FWD RTN RTN A2D CPU D2A CPE And beyond that… Future CPEs willrequires allow RFDS Spectrum “re-use” removing Coax from the equation Today the Coax and US to occupyby different portions of the Spectrum Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. What the Future Holds Copyright 2015 ARRIS Enterprises, Inc. All rights reserved. ARRIS–Confidential & Proprietary 30 1G Node-split to 64 subs/SG Node-split to 128 subs/SG D3.1 DS Limit = 10.8 Gbps (1200 MHz) D3.0 DS Limit = 4.9 Gbps (750 MHz) 10G DS BW for Modems (bps) 1.2 GHz of DOCSIS 3.1 Chans. 116 DOCSIS 3.0 Chans. 64 DOCSIS 3.0 Chans. 32 DOCSIS 3.0 Chans. 16 DOCSIS 3.0 Chans. 8 DOCSIS 3.0 Chans. 4 DOCSIS 3.0 Chans. 2 DOCSIS 3.0 Chans. 1 DOCSIS 3.0 Chan. Node-split to 256 subs/SG 100G Node-split to 64 subs/SG DS BW as a function of time (w/ ~50% Annual Growth Rate) Node-split to 256 subs/SG (De-Comb) Node-split to 128 subs/SG What will MSO's do with their HFC plants? Agg BW/SG 512 Subs/SG (2x 512 HHP Nodes) 256 Subs/SG (512 HHP Nodes) 128… 64… 32 16 Subs/SG (32 HHP Nodes) ~100 Gbps in 2030 100M 10M 16000 Subs/SG (Many Nodes) 8000 Subs/SG (Many Nodes) 4000 Subs/SG (Many Nodes) 2000 Subs/SG (Many Nodes) 1000 Subs/SG 512 Subs/SG 1M 100k 10k Agg BW/SG Max BW/sub 1k Nielsen’s Law’s 100 Tmax 300 bps in 1982 ~150 kbps in 1997 ~332 Mbps Agg BW/SG ~30 in 2030 Mbps Avg BW/sub in 2010 ~100 kbps in 2010 ~500 bps in 1997 10 1 1982 1986 1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030 Year Proposed Human Factors Formula: Required SG Capacity = K *Adv_Billboard_BW + # subs * Avg BW/sub Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 31 Looking Beyond the Top Billboard Tier: Example Residential Service Tiers % of Subs Tmax (Mbps) Tmax CAGR Tavg (Mbps) Tavg CAGR Top Tier – Billboard rate 1% 300 50% 0.48 41% Performance Tier 14% 75 32% 0.48 41% Common Tier 65% 25 26% 5%@1.92 60%@0.48 41% Economy Tier 20% 5 15% 0.12 20% • Based on study with multiple 100,000 Downstream Tiers - Tmax 10,000 MSOs • Strategy: – As each Tier hits 10G ceiling, peel off subs to NG-FTTP as needed Mbps Tier 1 Tmax 1,000 Tier 2 Tmax Tier 3 Tmax 100 Tier 5 Tmax 10 1 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033 Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 32 How Long will D3.1 & 10 Gbps Last? Example – 128 Subs per SG DOCSIS 3.1 Tmax Dominates Tavg Dominates DOCSIS 3.0 • Key Events: – 2024 – Top Tier moves to NG-FTTP – 2029 – Performance Tier moves to NG-FTTP – 2031-34 – Common Tier starts migrating • Tmax dominates through 2023 • Tavg dominates after 2028 Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 33 How Long will D3.1 & 10 Gbps Last? Example – 32/64/128/256 Subs per SG Tmax Dominates Tavg Dominates Tmax Dominates Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 Tavg 34 Residential Traffic Engineering Considerations • Tmax dominates in Near Term – Focus on increasing spectrum (e.g. 1.2GHz) rather than SG splits • However, Fiber Deep is also a tool to reach 1.2GHz – Maximize DOCSIS 3.1 capacity • Tavg dominates in Long Term – Moving Top Service Tiers to FTTP buys much time for majority of HFC subs – Gradual migration to Fiber Deep as part of Business As Usual to prepare for this era Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 35 FTTP Transformation: EPON or RFoG? FTTP Transition for Top Service Tiers to Extend HFC Life Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. FTTP Transformation: EPON or RFoG? FTTP Transition for Top Service Tiers to Extend HFC Life • Answer: BOTH!! • Use will have industry leading 10G EPON solutions – PFM (PON Fiber Module) for E6000: • • • • • Industry Leading OLT density 10G/10G and 10G/1G co-existence Built in Carrier Grade Redundancy for Optics Supports 256 ONU, 2K LLID per port Leverage installed E6000 customer base – Fiber Link Module (FLM) for PON Distribution: • Support Long Distances AND Large Fan-out – Support 256 ONU • OEO enables BOTH RFoG AND PON – Simultaneous support for both RFoG and PON • Leverage the best of D3.1 and EPON/GPON • Provide Best of Breed technology for each application Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved. 10 March 2015 ARRIS Confidential and Restricted For AT Academy 2015 37 Thank You!