Post on 09-Jan-2022
© 2008 Marvin A. Sirbu 1
CarnegieMellon
FTTP Networks: Topology and Competition
Marvin A. SirbuDepartment of Engineering and Public Policy
Carnegie Mellon University
sirbu@cmu.eduhttp://www.andrew.cmu.edu/user/sirbu/
© 2008 Marvin A. Sirbu 2
CarnegieMellon Conclusions Up Front
FTTP networks have significant economies of scalefacilities-based competition is unlikely to be
sustainableService-level competition can exist over shared network infrastructure
Sharing possible at different levelsSharing of dark fiber requires attention to fiber layout
There is great variety in the models of sharing which can be found today
© 2008 Marvin A. Sirbu 3
CarnegieMellon Outline
Models of Competition in FTTPAlternative FTTP architectures: impact on competitionEconomics of FTTP
© 2008 Marvin A. Sirbu 4
CarnegieMellon Outline
Models of Competition in FTTPAlternative FTTP architectures: impact on competitionEconomics of FTTP
© 2008 Marvin A. Sirbu 5
CarnegieMellon
Facilities based competition – each competitor builds FTTP network
Central Offices
ServiceProvider A
ServiceProvider B
Home 2
Home 1
Data Link Layer EquipmentATM, Gigabit Ethernet, SONET
Separate Networks
Network 1
Network 2
© 2008 Marvin A. Sirbu 6
CarnegieMellon UNE based Competition in FTTP
Dark fiber based – network owner wholesales dark fiberWavelength based – network owner wholesales wavelengths
Central Office
ServiceProvider A
ServiceProvider B
Home 2
Home 1
Data Link Layer EquipmentNetwork
© 2008 Marvin A. Sirbu 7
CarnegieMellon
Open Access based competition – network owner wholesales transport capacity
Central Office
ServiceProvider A
ServiceProvider B
Home B
Home A
Common Data LinkLayer Equipment
Network
© 2008 Marvin A. Sirbu 8
CarnegieMellon
Sharing Network Infrastructure: Summary
Layer: Shared Infrastructure…
0 Conduit and collocation facilities.
1 (Physical Layer Unbundling)
Dark fiber leasing, or perhaps, Optical Layerunbundling (CWDM or DWDM in PONs)
2 (Data Link Layer Unbundling)
Dark fiber and link-layer electronics at eachend. For example, Ethernet-based VLAN,or ATM-based PVCs.
3 (Network Layer Unbundling)
Basic network service provided. Forexample, IP Layer 3 service over cableusing policy-based routing to multipleISPs
© 2008 Marvin A. Sirbu 9
CarnegieMellon Outline
Models of Competition in FTTPAlternative FTTP architectures: impact on competitionEconomics of FTTP
© 2008 Marvin A. Sirbu 10
CarnegieMellon Home Run Architecture
Central Office
Equipment
OLT Port
ONU Central Office Infrastructure
Dedicated fiber to each Home
Feeder Loop Distribution
Loop
ONU Optical Network Unit OLT Optical Line Termination
Implications for Competition
Physical layer unbundling possible– wholesaler can sell individual fiber
Also supports open access
© 2008 Marvin A. Sirbu 11
CarnegieMellon
Central Office
EquipmentOLT
ONU Central Office Infrastructure
Shared Feeder fiber
Feeder Loop Distribution
Loop
N
1
Remote Node with Active Electronics Equipment
Active Star Architecture
Implications for Competition
Physical layer unbundling is difficult
requires competitors to collocate electronics at remote nodeMust provide feeder fibers for each competitor
Logical layer unbundling possible - supports open access
© 2008 Marvin A. Sirbu 12
CarnegieMellon Curb side Passive Star Architecture (PON)
Central Office
Equipment OLT
ONU Central Office Infrastructure
Shared Feeder fiber
Feeder Loop Distribution
Loop
32
1
Curbside Passive Splitter – Combiner
λ
λ
λ
λ
λ
λ
Implications for Competition
Physical layer unbundling not possible
Logical layer unbundling possible - supports open access
Separate λ’s may be used forData and video
© 2008 Marvin A. Sirbu 13
CarnegieMellon WDM PON
λ5
λ3
λ1
λ2
λ4
CentralOffice
EquipmentOLT
ONUCentral Office Infrastructure
Shared Feeder fiber
Feeder LoopDistribution
Loop
32
1
Passive Splitter –Combiner
λ1, λ2, λ3, λ4, λ5 ... λ32
Implications for Competition
Physical layer unbundling not possible Optical layer unbundling possible –wholesaler can sell wavelengthsAlso supports open access
© 2008 Marvin A. Sirbu 14
CarnegieMellon
Design Considerations in a PON: A Curb-side PON
Splitter 1
Splitter 2
Central OfficeOLT Equipment
Central Office Infrastructure
PON1
PON2
Neighborhood 2
Neighborhood 1
Central OfficeOLT Equipment
Both OLTs needed if only one home in each splitter group subscribes
© 2008 Marvin A. Sirbu 15
CarnegieMellon Design Considerations in a PON:
A Fiber Aggregation Point (FAP) PON
Central OfficeOLT Equipment
Central Office Infrastructure
Neighborhood 2
Neighborhood 1
Splitter 2
Splitter 1
Aggregation
Fiber Aggregation Point PON supports all models of competition
© 2008 Marvin A. Sirbu 16
CarnegieMellon
How many homes should be aggregated at an Optimal FAP?
OFAP allows deferring investment in OLTs until penetration requires it
© 2008 Marvin A. Sirbu 17
CarnegieMellon
OFAP as a Real Option to Phase-in New Technologies
GPON CO OLTEquipment
Central Office Infrastructure
Neighborhood 2
Neighborhood 1
Splitter 1BPON
Splitter 1GPON
BPON CO OLTEquipment
Aggregation•OFAP also supports flexibilityin future split ratios- 10 Gbps GPON, GEPON- WDM PONs
© 2008 Marvin A. Sirbu 18
CarnegieMellon
OFAP Benefits withan Active Star Architecture
•Higher utilization of RT and OLT ports•Neighboring homes can be served by different technology generations
Central Office OLT Equipment
Central Office Infrastructure
Neighborhood 2
Neighborhood 1
Feeder 2
Feeder 1
Central Office OLT Equipment
Aggregation point
RT & OLT tobe deployed as needed
•Larger serving area
© 2008 Marvin A. Sirbu 19
CarnegieMellon Sharing in the “Second Mile”
As video becomes dominated by unicast Video on Demand (VOD) metro aggregation network costs soarIn smaller communities, access to regional transport to a Tier 1 ISP is a major barrier to entryRetail service providers sharing an FTTH access network may also need to share at the metro/regional level in order to be economically viable.There is a tradeoff with distributed video servers
Sharing a content delivery network (e.g. Akamai) may be an alternative.
– This requires distributed colo space and interconnection
See Han, S. et al “IPTV Transport Architecture Alternatives and Economic Considerations,” IEEE Comm Mag, Feb 2008Lamb L., forthcoming.NSP, “A Business Case Comparison of Carrier Ethernet Designs for Triple Play Networks,”
© 2008 Marvin A. Sirbu 20
CarnegieMellon Regulatory Implications
If regulators want to be able to require dark fiber unbundling, they need to require compatible fiber layout
OFAP PON vs curb-side PONEven larger OFAP for competitive active star
– Need for additional feeder fibersAll architectures support logical layer (“bitstream”) unbundling
IPTV unbundling possible at bitstream layerIf video distributed over a separate wavelength, issues of access to RF multiplex.
© 2008 Marvin A. Sirbu 21
CarnegieMellon Outline
Models of Competition in FTTPAlternative FTTP architectures: impact on competitionEconomics of FTTP
© 2008 Marvin A. Sirbu 22
CarnegieMellon
Simple FTTH Economics: FTTH Includes Fixed Plus Variable Costs
e.g. for Verizon YE06Fixed=$850Variable=$880
Source: http://investor.verizon.com/news/20060927/20060927.pdf
$
Take Rate(R = customers / homes passed)
100%0%
Fixed costs
Cost = Fixed + R * Variable
Adapted from Friogo, et.al.http://ieeexplore.ieee.org/iel5/35/29269/01321382.pdf
© 2008 Marvin A. Sirbu 23
CarnegieMellon Cost Per Subscriber vs Take Rate
0
1000
2000
3000
4000
5000
6000
7000
8000
0% 20% 40% 60% 80% 100%Take Rate
Variable Cost Total Cost/Sub
$1730
© 2008 Marvin A. Sirbu 24
CarnegieMellon How Much Revenue to Support FTTH?
One operator estimates $90/month per subscriber$40 for ongoing services cost
$50/month to cover capital costsAssume an average of 10 year lifetime, 5% cost of capital
Fiber lasts 40 yearsElectronics lasts five years
$50/month can amortize $4700What if Average Revenue Per User (ARPU) is less?$30/month can amortize $2800
© 2008 Marvin A. Sirbu 25
CarnegieMellon Cost Per Subscriber vs Take Rate
0
1000
2000
3000
4000
5000
6000
7000
8000
0 0.2 0.4 0.6 0.8 1
Take Rate
Cos
t/Sub
scrib
er
Variable Cost Total Cost/Sub Capital at $50/mo Capital at $30/mo
Percent take rate needed to break even
Capital that can be amortized with $50/mo/sub
Capital at $30/mo/sub
Adapted from Frigo et. al.
© 2008 Marvin A. Sirbu 26
CarnegieMellon Cost Per Subscriber vs Take Rate
0
1000
2000
3000
4000
5000
6000
7000
8000
0 0.2 0.4 0.6 0.8 1
Take Rate
Cos
t/Sub
scrib
er Capital that can be amortized with $50/mo/sub
Adapted from Frigo et. al.
Consumers
Competition
Take Rate
© 2008 Marvin A. Sirbu 27
CarnegieMellon Economic Implications:
If revenue available to amortize plant is only $30/month, must reach penetration of > 45%
room for at most 2 facilities-based providersThis analysis understates the problem
No customer acquisition (marketing/sales) cost included– Customer acquisition drives up Fixed costs pushing
breakeven penetration higherUnlikely to see >90% total penetration
© 2008 Marvin A. Sirbu 28
CarnegieMellon Regulatory Implications
Facilities-based competition among fiber network providers is unlikely
Economies of scaleRegulators should be cautious of waiving open access requirements in return for investment in fiber
Could lead to remonopolizationAt best duopoly competition
If service competition limited to ISPs which own facilities greatly reduced service level competition
Operators will have Significant Market Power (SMP)Reduced service-level competition raises Network Neutrality issue
© 2008 Marvin A. Sirbu 29
CarnegieMellon
Central Offices
Service Provider A
Service Provider B
Home 2
Home 1
Data Link Layer Equipment
Separate Networks
Network 1
Network 2
Net Neutrality
Can third parties compete with vertically Integrated ISPs?
Apps+
Con-tent
Apps+
Con-tent
Apps+
Con-tent
© 2008 Marvin A. Sirbu 30
CarnegieMellon Conclusion
What are the different models of competition in FTTP? Facilities basedService level (over shared network infrastructure)
Fiber layout affects options for competitionOFAP supports fiber unbundling even for PONsMore feeder fibers required for competition
FTTP networks have significant economies of scaleUnlikely to support multiple facilities-based providers“Second Mile” sharing also important