future hfc directions: of i ccap, r phy and r macphy€¦ · future hfc directions: the role of...
TRANSCRIPT
Future HFC Directions:The Role of I‐CCAP, R‐PHY and R‐MACPHYJohn Ulm, Engineering FellowARRIS CTO‐ Network Solutions Team ANGACOM 08 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Bifurcation of the MSO CCAP WorldA More Complete Taxonomy of MSO Access Architectures
ARRIS @ ANGACOM 2016
Distributed AccessArchitectures (DAAs)
RemotexPON
Remote Pt‐to‐PtEthernet Switch
RemoteWiFi
Distributed CCAPArchitectures (DCAs)
Remote LiFi
OtherRemoteVariants
RemoteMACPHYw/o EQAMIn Node
Remote PHY
RemoteMACPHY
RemoteMACPHYw/ EQAMIn Node
OtherRemoteMACPHYVariants
RemotePHY
w/o EQAMIn CCAP Core
RemotePHY
w/ EQAMIn CCAP Core
OtherRemotePHY
Variants
MSO AccessArchitectures (MAAs)
Centralized AccessArchitectures (CAAs)
Head‐endI‐CCAP
Head‐endxPON
Head‐endEthernet
Centralized or Distributed?And Distributed Options are Still Evolving:• Remote PHY• Remote MACPHY (simple)
• Remote MACPHY with EQAM core• Remote CCAP• Remote CCAP with Centralized Controller• Remote OLT
208 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Do We Need Distributed Access (DAA)?Assessment Criteria for CAA/DAA Comparison
ARRIS @ ANGACOM 2016
Impact of spectrum and/or # of amplifiers
Impact of Fiber distance to node
Impact of the number of wavelengths
Facility/Node Space & Power Requirements
Any significant cost differences?
D3.1 Performance Impacts
Space and Power Utilization Impact
Cost Assessment (OPEX and CAPEX)
1
2
4
Impact on MSO OperationsSystem and Operational Complexity Impact 3
308 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
D3.1 Performance ImpactsDAA Digital Optics vs. CAA AM Optics
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
D3.1 Performance Impacts at ModemDAA Digital Optics vs. CAA AM Optics• DAA with Digital Optics
– Best overall performance– Same for all Distances– Performance still varies with Amp cascade
• CAA with AM Optics– Varies with Distance, wavelengths, Amp cascade
– 25km Performance almost equal to DAA Digital
– 80km Performance Relatively Poor• Drop Cable and In‐Home Impact
– Poor grade drop potentially wipes out all DAA gains
– Best results with Gateway @ point of entry• Isolate Home Network from HFC
ARRIS @ ANGACOM 2016
30
35
40
45
50
CAA ‐ 80km ‐ 16λ
CAA ‐ 80km ‐ 8λ
CAA ‐ 80km ‐ 4λ
CAA ‐ 80km ‐ 1λ
CAA ‐ 40km ‐ 16λ
CAA ‐ 40km ‐ 8λ
CAA ‐ 40km ‐ 4λ
CAA ‐ 40km ‐ 1λ
CAA ‐ 25km ‐ 16λ
CAA ‐ 25km ‐ 8λ
CAA ‐ 25km ‐ 4λ
CAA ‐ 25km ‐ 1λ
CAA ‐ 15km ‐ 16λ
CAA ‐ 15km ‐ 8λ
CAA ‐ 15km ‐ 4λ
CAA ‐ 15km ‐ 1λ
DAA – Remote Gadget 44λ +
0 Amp
1 Amp
2 Amps
3 Amps
4 Amps
5 Amps
6 Amps
1002 MHz Full Spectrum CNR Es mate to the RF Connector of Cable Modem at Point of Entry Through 1 Spli er
1024QAM
(10 b/s/Hz)
2048QAM
(11 b/s/H
z)4096QAM
(12 b/s/Hz)
8192QAM
(13 b/s/Hz)
16384QAM
(14 b/s/Hz)
512QAM
(9 b/s/Hz)
30.5
45
41
37
34
49
CAA to Node at 80km CAA to Node at 40km CAA to Node at 25km CAA to Node at 15km DAA at 80km+
Estimated Percentage of DAA Improvement vs. CAA
Estimated CNR
20% ‐ 30%
DAA has 0% to 9% Gain Against CAA
DAA has 9% to 18% Gain
508 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
CAA v DAAHeadend Space & Power Considerations
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Headend Space over the Years
ARRIS @ ANGACOM 2016
2012 Headend – 306 SG in ~46 racks 2014 Headend – 144 nodes, 112 SG
in 4 racks (28 SG per rack)
708 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
ARRIS CHP RX Optic SavingsD3.1 & PNM
Savings
D3.1 & PNM Savings
D3.1 & PNM Savings
PathTrak 1500
PathTrak HOU‐200PathTrak HOU‐200
PathTrak RSAM
PathTrak SWX
Sweep Return (JDSU)
Sweep Return RF Combining (ATX)
Return RF Splitter/Combining (ATX)
Return RF Splitter/Combining (ATX)
Return RF Splitter/Combining (ATX)
Legacy STB Savings
Legacy STB Savings
E6000 DCAM‐2
Broadcast Savings
Broadcast Savings
Broadcast RF Combining (ATX)
ARPD RF Combining (ATX)
Juniper 100G Switch – OOB Manage
OM2000OM2000OM2000
ARPDARPDARPDARPDARPDARPDARPDARPDARPDARPD
PathTrak HSM‐1000PathTrak HSM‐1000
Broadcast RF Combining (ATX)
Broadcast RF Combining (ATX)
Broadcast RF Combining (ATX)
Broadcast RF Combining (ATX)
E6000 IEQ Savings E6000 IEQ Savings
E6000 IEQ Savings
I‐CCAP Space Considerations – 2016‐17Upcoming Space Density Improvements
08 June 2016 ARRIS @ ANGACOM 2016 8
Power Distribution ‐ Fuses
Power Distribution ‐ Fuses
Power Distribution ‐ Fuses Power Distribution ‐ Fuses
Optical Cable ManagementFiber Patch Panel
I‐CCAP (E6000™)
Power Distribution ‐ Fuses
Optical Cable ManagementFiber Patch Panel
I‐CCAP (E6000™)
Aurora TX Optic Savings
Aurora TX Optic Savings
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – RX Digital Return Optics(Harmonic)
HFC – RX Digital Return Optics(Harmonic)
HFC – RX Digital Return Optics(Harmonic)
HFC – RX Digital Return Optics(Harmonic)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
HFC – TX AM Optics(Aurora)
Harmonic NSG 9000 40G
Harmonic NSG 9000 40G
Harmonic NSG 9000 40G
Harmonic NSG 9000 40G
Harmonic NSG 9000 40G
Harmonic NSG 9000 40G
Narrowcast RF Combining (ATX) Narrowcast RF Combining (ATX)
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
D3.1 & PNM Savings
D3.1 & PNM Savings
D3.1 & PNM Savings
Legacy STB Savings
Legacy STB Savings
E6000 DCAM‐2
Broadcast Savings
ARRIS CHP RX Optic Savings
E6000 IEQ Savings E6000 IEQ Savings
E6000 IEQ Savings
I‐CCAP Space Considerations – ARRIS 2016‐17192 Nodes, 192 DS/192 US HSD Service Groups; 192 VOD SG
08 June 2016 ARRIS @ ANGACOM 2016 9
Power Distribution ‐ Fuses Power Distribution ‐ Fuses Power Distribution ‐ Fuses
Optical Cable ManagementFiber Patch Panel
I‐CCAP (E6000™)
Power Distribution ‐ Fuses
Optical Cable ManagementFiber Patch Panel
I‐CCAP (E6000™)
96 SG per ⅔ Rack
Broadcast Savings
Aurora TX Optic Savings
Aurora TX Optic Savings
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – RX AM Optics(ARRIS CHP)
HFC – TX AM Optics(ARRIS Aurora)
HFC – TX AM Optics(ARRIS Aurora)
HFC – TX AM Optics(ARRIS Aurora)
HFC – TX AM Optics(ARRIS Aurora)
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
I‐CCAP Migration Summary:Space, Capacity & Power Case Study
ARRIS @ ANGACOM 2016
– ‘Chan’ equivalent = 38 Mbps
POD Migration 2014 2015 2015 2016‐17 2016‐17
Technology E6000 Gen 1,NSG UEQAM
E6000 Gen 1with IEQ
E6000 Gen 1IEQ
E6000 Gen 2 Today’s CHP
E6000 Gen 2 CHP RX/Aurora TX
Nodes per POD (4 racks) 144 144 216 216 480
AM Wavelengths 1 1 1‐2 1‐2 3‐4
SG per POD (4 racks) 112 112 168 192 480
SG per Rack 28 32 42 96 128
Subs per SG 320 320 213 187 75
Avail DS Capacity per SG 30 + 1x96 30 + 1x96 30 + 1x96 30 + 2x192 30 + 2x192
Avail Mbps per Sub 6.3 6.3 9.4 25.7 64
kW per rack 4.17 kW 3.0 kW 3.58 kW 3.08 kW 7.70 kW
W per SG 149 107 99.4 64 64
W per ‘Chan’ equiv 2.66 2.03 1.89 0.51 0.51
– While staying within existing power footprint!!!1008 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Headend Space Migration:2014 to ~2017 with I‐CCAP, R‐PHY & R‐CCAP
ARRIS @ ANGACOM 2016
Existing Headend2014
I‐CCAP Headend~2017
R‐PHY~2017
Access Architecture (Central ‐ CAA or Distributed ‐ DAA)
Time Frame
Space NeededFor ~200 SG
SG per Rack SG Scale
Older CMTS + Low Density EQAM + Older AM Optics Pre‐2014 20‐50 Racks 4‐10 SG 0.2‐0.4X
CAA E6000 + High Density UEQAM + Ext B‐cast + AM Optics 2014 ~7 Racks ~28 SG 1X
CAA E6000 Gen 2 with B‐cast + AM Optics ~2016‐17 ~1½ Racks ~128 SG 4.5XR‐PHY DAA: Gen2 MAC Core + Ether Aggregation + WDM ~2016‐17 ~⅔ Rack ~288 SG 10XR‐MACPHY DAA: Router + Aggregation + WDM mux/demux ~2016‐17 ~½ Rack ~384 SG 14XEarly R‐CCAP DAA: Ethernet Aggregation + WDM mux/demux ~2016‐17 ~⅟5 Rack ~960 SG 34X
R‐CCAP~2017
R‐MACPHY~2017
1108 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Distributed Architecture Comparisons:R‐PHY, R‐MACPHY, R‐CCAP
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
A More Complete Taxonomy of MSO Access Architectures
ARRIS @ ANGACOM 2016
Distributed AccessArchitectures (DAAs)
RemotexPON
Remote Pt‐to‐PtEthernet Switch
RemoteWiFi
Distributed CCAPArchitectures (DCAs)
Remote LiFi
OtherRemoteVariants
RemoteMACPHYw/o EQAMIn Node
Remote PHY
RemoteMACPHY
RemoteMACPHYw/ EQAMIn Node
OtherRemoteMACPHYVariants
RemotePHY
w/o EQAMIn CCAP Core
RemotePHY
w/ EQAMIn CCAP Core
OtherRemotePHY
Variants
MSO AccessArchitectures (MAAs)
Centralized AccessArchitectures (CAAs)
Head‐endI‐CCAP
Head‐endxPON
Head‐endEthernet
1308 June 2016
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Distributed CCAP Options – Pick Your Favorite
08 June 2016 ARRIS @ ANGACOM 2016 14
RMD – Minimal Config
RMD with Embedded EQAMRemote CCAP with Centralized Controller
RPHY System Components
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
DAA Topology Migration Example –Multiple Options• DAA Shelf at Hub or Cabinet Location• DAA at Existing Node Location• DAA at Fiber Deep Node
– Might need 12‐24 RPD/RMD per existing Serving Area• Each might require its own Long Distance 10G Optics too
– Fiber Deep node may have ~60 HP with only~30 subs• Combine 4+ RPD into SG in MAC Core
• Hybrid DAA with AM Fiber Deep Nodes– Shared RPD/RMD resources– Lower cost, lower power Fiber Deep nodes– Scale RPD/RMD from 1x1 to 1x2, 2x2, 4x4, . . .
08 June 2016 ARRIS @ ANGACOM 2016 15
RPD/ RMD
RPD/ RMD
Copyright 2015 – ARRIS Enterprises, Inc. All rights reserved.
Future HFC Directions:The Role of I‐CCAP, R‐PHY and R‐MACPHY• Centralized or Distributed Architecture
– Centralized CAA Headends (I‐CCAP + Optic Shelves) have plenty of Life left in them for SG Growth• I‐CCAP + Optic Shelf improvements handle most space/power growth needs• D3.1 Performance is comparable for less than 40km• DAA makes sense for very Long Distances, e.g. 80km; and for extreme Headend Space consolidations
• Distributed Architecture Options– R‐PHY, R‐MACPHY and R‐CCAP have various Pro’s & Con’s
• R‐PHY provides minimal size/cost/power in node; critical for smaller nodes as common in Europe– Multi‐vendor interoperability
• R‐CCAP with Centralized Controller provides optimal Headend Space consolidation• R‐MACPHY pushes a middle ground between these
– DAA Topology Options• DAA Shelf in Hub or Cabinet does not disrupt existing HFC plant while providing benefits of Digital Optics• DAA in Existing Node Location is simple step; But still have Performance degradation due to cascade Length• DAA at Fiber Node Location provides optimal performance, but at a cost• Hybrid Shared DAA with traditional AM Fiber Deep nodes provides the best of both worlds
08 June 2016 ARRIS @ ANGACOM 2016 16
Thank you