TIlE COST CONVERTING UTILITY NETWORK INTO A NETWORK
BROADBAND AND CABLE TELEVISION TO SUBSCRIBERS
October 18, 1990
Report to The National Regulatory Research Institute
Prepared by
Julia A Miezejeski Michael Miller Bruce L. Egan
Northern Business Information!Datapro a Division of McGraw~ Hill, Inc.
157 Chambers Street New York, N.Y. 10007-1015
This report was prepared by Northern Business Information/Datapro for The N adonal Regulatory Research Institute through funding provided bl participatin~ fllember commissions the National Association of Regulatory Utility CommiSSIoners (NARUC) and a grant from The Center for Advanced Study in Telecommunications (CAST) at The Ohio State University. The views, opinions, and costing models of the authors do not necessarily state or reflect the views, opinions, or policies of the NRRI, NARUC, or NARUC nlember commissions, or CAST.
3
· . 1
3 1
4
5
6
7
. 81
L
1.11
1
to
7
6.1
7
v
4
5
, ... 6
8
9
. 11
... 49
.51
public policy over whether should allowed to provide cable television services has largely centered on cross ownership and dominance issues. However, another important issue, the cost converting a public switched telecommunications netw'ork one of delivering the full
of broadband and cable television services, has less attention. Because of the variety of the broadband network design options available to
telephone utilities, state commissions have a need for a reliable framework that can to identify the broadband deployment option being implemented by a utility
and the cost of the o:ption. This report by Northern Business InformationjDatapro will allow a cOITnmSSl0n to identify options and costs.
Of course, the NRRI, NARUC, or NARUC member states do not necessarily endorse the particular costing method employed in the course of this study. The NRRI feels, however, that the simple engIneering cost model used is a reasonable model and provides a benchmark that can be used by other cost studies. Further, the analysis herein does not necessarily assume or favor any particular ratemakin~ method. We appreciate the supplementary financial support for this project proVIded by the Center for Advanced Study in Telecommunications of'Ine OhlO State University.
vii
Douglas N e Jones Director NRRI Columbus, Ohio October 1990
cllstonlers on
new
or
high-speed
are
costs are
costs shared and switching plant traditional and nontraditional
services. Nevertheless, this is the challenge, and some
identifying classifying costs are contained in Chapter L
xi
1
;'1L(I>l"~<;Il~~n ........... 1!"""' ....... ~:!o·W'~1r·1"'n urith';·" traditional lines of business has the
exchange telephone companies to diversify seek new sources of revenue.
area draws considerable interest is the cable industrye However~ the
telcos face legal restrictions barring them from. operating cable TV networks.
Prohibitions the 1984 Cable Communications Policy Act prevented the
regional bell operating companies (RBOCs) from operating cable TV networks.
Recently, both federal and state regulators have begun to reconsider these
restrictions. The costs of new fiber~based networks capable of providing cable TV
services will playa vital role in the debate. Equally important are the costs
associated with other broadband services, apart from cable TV. Telcos are eager to
exploit opportunities selling other broadband servi.ces such as video-on-demand,
home banking, and home shopping. From an operations standpoint, once a
broadband network is in place, telcos would then package existing narrowband
services*-plain old telephone service (POTS) and low-speed data--together with the
broadband delivery to customers.
Traditionally, investment detisions by telecos and cable TV operators alike
were made in a monopoly environment dominated by issues of service quality, cost
savings, and capital recovery. Unlike investment decisions of the past, future
investment decisions will be made in an increasingly competitive environment.
No doubt, both LEes and large cable TV operators view aggressive deployment
of fiber as a strategic advantage. tlWhoever gets more fiber in the ground first,
wins!" seems to be the prevalent thought process in the telecom conununity.
Despite current regulatory and legal restrictions, telephone companies see that
installing high quality, high capacity fiber in the network will position them to meet
the future needs large customers and third-party vendors whose services
currently cannot be provided by telephone companies directly. What's more,
telephone companies believe that in the near future, fiber lines will be cheaper to
install and maintain, and will provide clearer connections for voice conversations.
Underlying these plans for generating new revenues and cutting operating costs is
1
dollars V'>.P9IiI"';.ft..:::. "'''fI:'1-iO.1i''n into a full
spent today's
Clearly, this
largest cou*;truction ever undertaken by the telco community.
telephone companies as a serious competitive Cable
cable television business. face a
local television monopolies. some telcos, cable
operators are on expanding and envision intercity video distribution using
fiber optics to provide the programming requirements for several local systems.1
TItis could enable cable television's penetration in the telecom business for
internode transport and ultimately to connect end-users. Cable TV firms also have
growth strategies that call for diversification through new service applications.
They are upgrading their networks with fiber in anticipation of this growth. If
cable television is first with ubiquitous fiber installations, widespread bypass of
telephone company local facilities will be likely.
A clear cost structure, segmented by service offerings and by network
configurations is needed to formulate responsible public policy for the telephone and
cable TV industries. For policy makers, the need for a generic costing mechanism
is urgent. Telcos and cable TV operators are now considering several strategic
plans for building broadband networks. These service providers have remained
flexible in their network planning. Recently, the RBOCs chief network planners at
Bell Communications Research (Bellcore) endorsed a residential broadband
architecture that is a radical departure from the architecture that BeHeore endorsed
previollsly.2 In this fluid environment, the purpose of this report is to develop a
framework to identify and analyze costs associated with the specific construction
alternatives that a LEe could pursue to provide cable TV services and other
broadband services.
The remainder of this chapter win provide an overvie\v of subscriber loop
architectures, network evolution and telephone company fiber loop trials.
1 "Kahn Plans All-Fiber Overbuild in N.J.," Cablevision, October 12, 1987.
2 Special Report SR-TSY-001681, Issue 1, June 1980.
2
Currently o;:I.LI'lb.""'''''.!UI. different subscriber loop architectures being
n.o."M"","",,,,varchitecture
will limit our descriptions to the
category.
' ........ "" ......... ..., a U ........... A ............
dominant architectures
Fiber-to-the~·home is characterized by use fiber-optic facilities (shared
and dedicated) in construction of the loop portion of the network to the
subscriber's home. The first FITH installations were deployed in traditionalltstarj,
architectures whereby service is provided on equipment dedicated to a single
customer. Telcos began using the star technology because it allowed them to use
the existing distribution equipment offered by suppliers and was consistent with
their installed base and operating and engineering expertise. For exanlple, AT&T's
switched*star architecture fits well with its SLC Series 5 subscriber loop carrier
system, which has been installed widely by all RBOCs since the early 1980s.
The single·star architecture uses dedicated fiber-optic facilities from the
central office all the way to the subscriber's home. The only resource that is
shared is the central office. Figure 1.3 shows the network components utilized in
a single star architecture.
!lQilhle-Star Ar~hitectu.re
When multiplexing equipment is introduced to the system at remote terminal
sites, the architecture becomes a double star. Some resource sharing is provided at
the remote terminal (the serving area jnterface or where optical signals are
3
~ Fiber
Feeder
I
Serving Area Interface
Remote Terminal
To CATV Provider
Fiber
Distribution
Optical Network Interface
Drop
cr (D ., i
-i a 9F-1 4~ =r\-l ro
9 :t: o 3 ro
Optical NAI'Ulftlrlif
Interface
(JI
Central Office
Fiber
Feeder
I
Serving Area Interface
Remote Terminal
LTO CATV Provider
Fiber
Distribution
COAX
Optical Network Interface
COAX
'" Copper Twisted Pair
~I':J 0" \-Iml.Q ...,~ II;
-1(1) o I -1~ ::r O
(1)1\.)
I n IC ..., C"
U)~ :;'/-1-
(OI...O --r:;: rot-; 1(1)
U)
(;1-' ,e III
I J;>w
In ito"t,. , 0 =r
s Optical Network Interface
Fiber 0"1
Features: All resources dedicated Optical NAtwnrill' ll!nt'ftlri'~iJ"I>_
Optical Network Interface s
I' 1 d mu. .. t1p exe
multiplexing
home.3 loop ...... #"iI·u'l-·""' ... ~·~~ ... ,""'''
usually "",VA..M.:IA";''',",
digital ""~"'_JI,.L""'A systems. Figure provides a simple of the double-star
(switched) Passive
a nevv technology currently being studied several
telephone of resource sharing a cost-effective
m.eans upgrading to broadband services. l-Iigher levels of resource sharing are
achieved when one channel becomes multiple shared channels lightwave
signal splitting. Additionalchannels can be allocated among subscribers as needed.4
This technique uses fiber splitters, optical couplers, and/or wave division
multiplexing (WDM). An example of passive optical networking is illustrated in
figure 1.5, which shows BeHeorc's passive photonic loop architecture. This
arrangement is capable of POTS and cable television-type service on a single fiber
optic access line. Wave division multiplexing techniques may be used to provide an
analog broadband service capability while at the same time providing narrowband
digital or analog POTS. WDM uses electronics to "channelize" a single fiber loop,
allocating bandwidth withinthe same physical facility to each type of customer
service as required. This architecture is also referred to a,s, a passive (nonswitched)
double star .
. Fiber to the Curb
Fiber-to=the-curb architectures are characterized by the presence of fiber
facilities from the central office to the pedestal. With F1TC, the pedestal is the
point at which the fiber is terminated and the copper or coaxial cable begins
distribu tion.
The "last mile" (slang for the last segment or dedicated subscriber loop
portion) of distribution cable, opto~electrorJc conversion 'and subscriber electronics
represent the major costs associated with FITH. By contrast, with F1TC, opto
electronics are shared at the pedestal by at least four homes (rather than
dedicated to each individual subscriber as in FITH). In addition, F1TC drop cables
3 Mary Henry and Daniel F. Zinsser, liThe Telescope: Telecommunications Equipment Trends and Directions," Goldman Sachs Research (March 1990)~ 9.
4 Ibid.,17~19.
7
00
le ... Star
Serving Area Interface
Remote Terminal
II ~
Serving Area Interface
Remote Terminal __
Features: Some sharing of facilites occurs
remote terminal
Fiber
Or:.:! 01-'-CI.Q
~~ Wr; am
tn """'"j--l (» • .,~
II rc I r+ m n r+ c: ., CD
~
Optical Network
Fiber
Fiber
Optical Network
1I_\li'1_~1I Network Illn1'Qn:::r.ra
1..0
Ii
I ,.
Serving Area Interface
D
Remote Terminal
Assumptions 16 Subscribers per remote node Services include:: POTS, 2-WclY data and
video conference, i-way 4 channel video and video database
* SeUcors Passive Photonic loop
Cost (per Subscriber) Serving Electronics OlE, E/O WDM Drop and Distribution Feeder Other Total
-
$2,000 700 250 400 200 300
$3,850
the pedestal to the home are copper or coaxial
most popular
UI1JU}-Snu and bus.
Triple-Star Architecture
architectures in """A.''''V~ .. '''V'''''C'''' company
1l"<IJIT,n,Q1l" than fiber.
today are the
This FITC system employs a triple h stal' architecture where fiber runs to the
pedestal and standard twisted pair copper cable runs from the pedestal to as mfu"1Y
as four homes.5 Certain phone companies view architecture as most beneficial
for rehabilitation projects and short loop constructions. The increase in resource
sharing in triple~star installations will drive the cost per subscriber down closer to
today's copper based installations. The t.riple-star architecture is depicted in
figure 1.6.
Bus Architecture
The bus architecture is the most popular (and economical) FITC system for
narrowband applications. The Raynet loop optical carrier (LOC) system is being
trialed by Ameritech, BellSouth, and NYNEX. One bus will support as many as 192
subscribers. However, original telephone company construction will support a lower
number to provide for growth and additions. Skepticism prevails in the industry
over whether this design can be easily upgraded to FITH or to two-way broadband
cap ability. 6 Figure 1.7 displays the bus architecture. The bus design was
originally developed for multimode fiber but Raynet's new LOC2 system. win accommodate single mode fiber, the current industry standard~ Raynet claims that
through passive optical networking, this new system will support up to 384
telephone subscribers and offers upgrade ability to broadband services.
Fiber BackbQn~N~twQrk] 'The hybrid fiber/copper or fiber/coax network generally implies a fiber~optic
backbone, or trunk, interconnected to either telephone company twisted pair or
cable company coaxial cable for the last network segment. The fiber backbone
topology will likely become the new cable TV industry's standard for implementing
5 Ibid., 9.
6 Ibid., 13-17.
10
· O"l
IilI
I II
I rc I Serving Area Interface
Remote Terminal
Serving Area Interface
r-a I Remote I Fiber /--l Fiber Terminal
Fiber
Optical Network Interface F· "
Pedestal
\.....I N
HemOte Terminal
Pedestal Optical Net""ork Interface
Serving Area Interface
Features: ] Up to 8 homes/pedestal
to 24 pedestalslbus .
Copper
r it t
~ ........... :£ ==(
Pedestal Optical Network Interface
Pedestal Optical Network Interface
Copper
Pedestal Optical Network Interface
Pedestal Optical Network Interface
.. ~~ \..;>!OIt)r~'!V" ~<.<>
eJlisting cable feature tree-and-branch, one
for bi
directional amplifiers. However, the tree-and-branch architecture's many coaxial
amplifiers cause system problems as signal errors cascade
through amplifiers. By using fiber instead the
the with are
""'A .................... ."."'."'~9 enhancing signal quality and service reliability.
fiber optic points interconnected thereby allowing for signal
redundancy or alternate routing possibilities in the event of node failure.
addition, a fiber backbone can enhance picture quality, expand overall system
bandwidth, and allow for two-way narrowband service possibilities at minimum
additional cost.
Telephone company fiber backbones (fiber feeder plant) are less interesting
from a customer's perspective since the service capabilities of copper loops are not
significantly enhanced for residential customers. The primary motivation for
telephone company fiber backbone deployment is the cost efficiency of high density,
shared plant. Basically, the telephone company fiber backbones are deployed to
replace nonfiber trunk and feeder plant. The cost justification logic for telco fiber
backbones is identical to that which justifies use of any replacement of copper
pairs: to expand capacity and narrowband signal quality and save on maintenance
costs.
Perhaps the most significant result of telephone company deployment of fiber
backbones is that cable companies or others may be able to efficiently interconnect
to it from headend or fiber-hub points in order to achieve intercity two~way
switched and point-to-point service. This situation is intuitively appealing to
industrial logicians since the relative strengths both of telephone companies (with
high bandwidth interoffice and intercity facilities) and cable companies (with high
bandwidth local distribution facilities) may be combined for the benefit of
residential subscribers. Figure 1.8 provides a basic view of the fiber backbone
architecture.
Loop Network Evolution
Now that we have provided general descriptions of the basic fiber loop
architectures, it is important understand these network architectures may be
13
Figure 1 .. 8 Fiber Bockbone Net work
14
evolve to a today's telephone
........ ""',' ............ , ... ..., of telephone
""'V .... .w.f-' .. ~ ...... ~"''''' and other communications providers are generally specific the
particular type These configurations are
digital
simultaneously a services. With broadband network technology, such as
transmission are commodityqlike, and customers win be able to use capabilities
for wbatever final services they demand. Ultimately, customers may be able
obtain a host of basic network functions over a single access facility.
The first stage of the network evolution, (assuming interexchange trunks are
already converted to fiber) begins with integrated services digital networks (ISDN).
ISDNs feature a single integrated access link including access to a host ISDN
network switch, and intelligent signaling network. Basic components of the
signaling network include ISDN switches, digital transport facilities and signaling
systems, including si~altrans~er points (STPs) and network control points;(NCPs)
for database services~,~;:
At this stage the copper feeder cable is the next logical segment to be
replaced by fiber, thereby creating a "fiber backbone" system that may evolve to
FITH and FITC systems~ using star, bus, or star-bus configurations. Remote nodes
(remote terminals) located on the customer side of the host ISDN processor (serving
area interface or SAl) may be used to provide;,~ert.ain features and functions. The
serving area interface becomes fiber-c~mpatible :ih~ough the placement of
appropriate digital loop carrier technology. The use of sophisticated equiplnent or
"intelligent" remote nodes may provide customers with "smart" access. The features
and functions which "smart" access provides will likely be in the category of
enhanced services, not POTS.
In stage 2, copper loop distribution cable may be replaced with fiber. Next, at
an interim stage 3, the costly opto-electronic (O/E) conversion function may be
installed at a pedestal and shared by many customers through FITC installations.
These installations require the placement of fiber~compatible pedestals, point
which optical-to-electrical and electrical-to~optical conversions take place.
The final migration is to stage 4 (FITH) as the subscriber
replaced with fiber the primary optical network interface may be TTln'u.o.n to the
subscriber's home. This last network segment (drop) may becorne longer than the
15
j-I 0')
Digital or Analog alOffice
,- -10 hn
f-! ~
1..0
__ Copper or Inter Exchange Trunks
".----- Copper t-eeaer t;CiI:Ues
Serving Interface
to ornl1nil"ll"'ll&l>'I"Ilt
pedestal Traditional
timing of this
technology and economics.
evolution.
depicts
customers from
loop network
In this C'A§'· ... ~n.n we will present brief descriptions of selected telephone
company fiber For a complete list trials, see figure 1.11.
Contel
Conters trial in Ridgecrest, California uses AT&T's FITH double-star
architecture. The rerrlote terminal (two SLC Series 5s) is located about three miles
from the Ridgecrest central office and provides 192 houses with two voice lines
each. At the central office an AT&T 5 ESS provides the switching for POTS
brought in to each home on fiber at a 1.544 Megabits per second (Mb / s) rate.
Since the goal of this trial \vas to make it as cost effective as possible, only a
single fiber is deployed to each house. A distant terminal (AT&T's name for its
optical network interface) is flush-mounted on the side of each house. Th~ distant
terminal is powered off the IOO-volt AC source provided to the home. Eight-hour
battery backup is provided. We esthnate that the customer's power bin would be
increased at most by $0.18 per month. Fiber cables with a maximum of ninety-six
fibers are used for the distribution loops.
Conters Sydney, New York FITH trial win be a rehabilitation job that will use
fiber cable strung from poles. Construction for this trial began in July 1989. One
hundred sixty-six residential customers, thirteen small businesses, and five major
businesses were scheduled for hookup by March 30th of 1990. This trial is one of
the first fiber optic aerial trials ever to be conducted. Initially only POTS will be
provided, but delivery of data and video will be considered later. In all, the
Sydney Lightwave Project construction costs are estimated at $1.2 million. This
trial utilizes the AT&T SLC series 5 system over single mode fiber. The
optical/ electrical unit in the distant terminal will use approximately 2.5 kilowatts,
costing ratepayers roughly $0.23 per month in extra power charges.
j-J 00
feeder
Stage 1
Serving Area Interface
Remote l'erminal
Replace with fiber 'Fiber Backbone"
r
Distribution
Stage 2 ace with
fiber
Stage
I iii
I
"New" pedestal (fiber to share or
o :l
FIGURE 1.11
TELEPHOl'i"E COMPANY F1BER LOOP TIUALS
Number
Start of of
Telco Location Date Homes Se:n.vice
AUtel Glen,NC 1989 50 POTS R-Tech FITC Ameritech Jefferoon Maws, OH 1990 100 POTS FITC Centel Tallahassee, FL 1989 100 POTS AT&T FlTH Contel CA 1989 100 POTS AT&T FTfH Conte! Sidney, NY 1989 600 POTS
Conte! Rancho Las Flores, CA 1991 350 POTS, video, other AT&T, R-TECH
CA enhanced services FITC Contel Wyoming,MN 1989 240 POTS AT&T FITH Bell Atlantic Loudon County, VA 1989 126 ]'OTS, digital video BBT
Bell Atlantic South Brunswick, NJ 1988 104 POTS and data AT&T FITH Bell Atlantic Perryopolis, PA 1989 100 POTS, CATV, switched PM ftJcatel FITH
video, multimode fiber
BeIiSouth. The Landings, GA 1989 192 POTS AT&T FITH BeHSouth Lakeview Terrace, SC 1990 100 POTS AT&T FITH BellSouth NOfC1'Q6S, GA 1990 N/A POTS Raynet FITC BeUSouth Sawgmss, FL 1990 N/A POTS Nfl FrTH BeUSouth Memphis, TN 1988 100 POTS AT&T FTTH BellSouth Coco Plum, FL 1989 200 POTS AT&T
:seUSouth. Governors Island, NC 1989 49 POTS AT&T
BellSouth Heathrow, FL 1988 4000 POTS, ISDN, digital CATV, Nl1 1'"TIH transport (data security,
meter reading, energy mgmt)
BeUSouth Hunter's Creek 2, FL 1989 117 POTS AT&T FITH BeUSouth Hunter's Creek 1, FL 1986 251 Digital CATV AT&T FITH BeUSouth Morrocroft, NC 1990 126 POTS AT&T FTI11 Cincinnati Bell Cincinnati,OH 1989 100 POTS AT&T Frn-I GTE . Cerritos, CA 1989 SOOO POTS, digital CATV, AT&T FITH
advanced broadband GTE GTE
(video on demand, home Amer.LW
banking, Arner. Lightwave
shopping, security,
utility meter reading)
:NYNEX Lynnfield, MA 1990 100 POTS FITC Southwestern Bell Mira Vista, TX 1989 100 POTS and CATV Amcr.LW FITH Southwestern Ben Leawood, KS 1988 134 POTS AT&T FITH Southwestern Ben Olathe, KS 1989 260 POTS AT&T FTTC US West Mendota Hghts, MN 1989 100 POTS AT&T FITH US West Scottsdale, AZ 1989 % POTS AT&T FTTH
.1lU'''''.luu. .... v'M-I,. ..... is deploying AT&T's SLC Series 5 Feature Package
as if Project Phoenix") for POTS only in FITH configurations ! ....... "",.. ......... locations:
> Hunter's Creek Orlando, Florida serving 117 homes
> Coco Plum, Miami, Florida serving 119 homes
> Landings, Savannah, Georgia serving 192 homes
> Governor's Island, North Carolina serving homes
> Morrocroft, North Carolina serving 96 homes
> Lakeview Terrace, Charleston, South Carolina serving 100 ..... v .... .!!..!!....,'"
> The Grove of Riveredge, Memphis, Tennessee serving 75 homes
Standard POTS channei units will be housed
channel unit provides two voice frequency channels per fiber.
the central office, RT, and the distant terminal (optical network interface) 9
VI.J ..... .a. u,'i."-' at a rate of 1.5 Mb / s. Bi-directional optical transmission is
accomplished at 1,300 nrn. The distant terminal which is a weather-proof
on the wall customer location, has of
The DT will run on ",u.,;:c;t·~"JA"'!'\\.ltf,'.l.'" ..... ~ ... A",.. .rio,
20
same
fl1?T""'lI"n Telecom's
Mb I s are to each subscriber to
video chamlels, one basic 144 Kb/s ISDN channel,
channels.
1I"Il1" .... u..::., ... v was initiated June 1988 with the installation of a single-mode
POTS service. In November, ISDN was added .
. ~"1II1!1"1't""'~·H::.d1! video was Service was initialized originally to 256
..... ""' .... ~.u ............... of 4,000 homes will eventually receive service.
'I.<I. ... ~ .... "'(!,I!, ... set-top-converter and wireless remote control unit. The
V1p.lJilP·r~ to can pay for video~on~demand by pressing a button.
system's upstrearrl signaling capability over
signal downstream the home. is
this corrrrnunity. l'he project cost is $3,001,245.
uses
an existing neighborhood.
"-".ll. " .... ' ...... ,"" ... covers an area length is 13,900 remote
to
environmental vault, J;..lt'U'I.<l-L:7""'.:7I
Raynet's
.... "'4Ibll-JlV.!LlLJI.""' .......... N\'NEX win most """ ......... C>.'il"~ ..... "''''' in 1991.
for ........... · .......... uh .. ' .....
production w VlL'~.l."'j\'"".:JI
21
costs
assume it
are 2
for categorization
""',; ......... ""', .... "' .... .., ... hroadband systems. The their
.. :U.JI. ..... ...,. ........ enough that any regulator or
the fundamental netw'ork architectures
to be general enough in specification that any into the generic rnodel.
illustrations the report show variations
differences in architecture and -n ..... 'i-;O'1i"'IIi-1
prospective loops.
architectures which occurred in
.I..3..I.'W.au.uJLI. ... .F. some which are only on the drawing board. it plant how the evolution of "H1I .... '~.,".
"::»UJm~""".:Jll. a logical n1l"£\Or.t.!!>CC'1inn
is obvious that digital
""' ............. """'c ........ ,""' ... subscriber ...."...,"" ......... ,"''''''"
all the way
a near-term one. ""' .. "' ... ~-"'I"M'''·.''..:'lIn1i4lC' will be concerned
......... "' ................. Jl.Jl..F! current l1ra.T~""'i".:.i"·.r'n
a pressing one
it next ... ""'· .. ·, ....... ·"'-'~'E~·n
22
is not a
hinges on subscriber ""''''''''~JL''
",V",''''='llA-A'\,,> """""#-",,, .. "" ...... """ assumptions; it a who will have
are
cases current SIa.le~UI-
costs are
are
>
>
to Iueasure
sources a.re
2
on
1.
3
customers are !f;';::"~Rr""',fl
are:
!>:"" •..• ", • ..", over is
as are
a
II
HOV
Voice & ;;;:~flll"ih1:.1lI'ff'M"il .& Overhead Bits
DS1 Voice Circuits
~---,."""""""'~
KF Bdstioo Fiber
~
I
Piaca 1.0KF of Fiber Cable
so in mass
2 1 1 1
costs
4
$
costs
> serves
3
:>
5
or
some
can
L
w III
I
#!i'mlll1l1Jl!ll)'i!I~~~'#.AA:w:x.~~ ~~~~~~~~
;::
~ ...J
Hvbrid Co
Assumptions 10,000 Subs 375 total miles of plant 51 miles of fiber backbone The fiber backbone is the
only part of the plant being replaced
* Based on ATe
Supertrunk (Fiber)
IFiber ystem*
Cost (per Sub):
-nt-:j C=l-Io co 1.0 ..,~ (Dr; oro o "U"I C" • o~ :J co :z co ,.... ~ o -, "
Electronics $17.9 Fiber Trunk 18.1 Total $36
Percent of total cost Electronics 49.6% Fiber Trunk 50.3%
sources
"""""u, .... ,,·.""" ""t/""""""'ii"'l>nif\"\(l !:'~·~Mli",""h.o,rl w"'L.:.h'1A1P'V ""~"-'L<~"''' ... " to ·,,,,,..,,, .. H7<:;1;
more,
4
cost data available .. _ ... -,~A{,;"wn1T,.a,,,.,. ... costs could
which was .".a. .... ""'''"',1I"h.1>"'>"<'''r ......... .n. ... .:.ri or. ...... """''''''''
see an
~ 1..0
Lt-:j m f-'o '1..0 'c ~t1 0.(0 r CJ)U1 CT • (J)U1
Jerrold Labs "Svstem " (f)
u:: fI) ,CD
Heaet Enel
Fiber Super Trunk
Approximate cost/sub $1000
Assumptions: Hub serves < "1000 subs. 128 total TV channels . 4 TV channels per home 20 digital audio channels Upstream data link 10 kbs 2-way voice and data
capable, up to T1 rate
Switch Site
(Hub)
Distribution Fibers
Each Fiber Serves 4 Homes
TAP Fiber/Coax
3 A
:
Tap Serves 4 Homes
Coaxiai Cable Drop
categories of generic model "."",,,,...,,,, .. ,,y some !troles that the Q"7~;1l"4J!(1,,;;l';
(";08t
> usually represent subscriber loop plant.
total cost
'\J"".lUi:lHl ....... Jl~ M. Jacobs Designers, tI
ff'>_.g.~'Jt~.""a IEEE; 1978, 387~390.
(channel selectors)5 mUJU[)le}j~ers optical receivers~ power, ,j,""-_ ....... '"
"'In·'I'<::>~'f'lI"""'" line cards, circuit packs, "'~:.r.·,-e ... 'ttre"i"' .. lC'
charger
conVf',rter 5.0
5.0
18.7
51
INVESTMENT COST PER HOME OF WITH ALTERNATIVE NETWORK ARCHITECTURES
(In
Network
loop c&rrier Nl.nowbMd on fi~r Voice and diatribu ted video IBN
IBN
IBN
Coaxial cable Fi~r backbone
Narrow. bemd
<It
..
Diltrih-uted Switched
Video
Activ~ Double Star
PaHive Double Star
Switched Star
Cable Networks
'" t1
52
Cent"'! OfficI!
~ .l
1 2
10
40
ONI
6
&moUl Ov~nUl Nooe ONI To~l
~t
---.~
36 37 45 150 196 60 250 3n 75 300 3""'" I I
300 :310
----.-----~. ------300 340
Trunk & FefKil!!'
17 17 is 21 ----------
cost.
> cost
costs is rIl'''·,...n:.:1>.r'I
>
>
>
>
>
*
*
*
>
>
6we
for
on
some our are
>
>
>
>
to
!.l..Ujt4lUJlUW,VJ.UI. costs n<:1i.<m"'ur.l!~lI" 6 d'''r>o",,,d',,,,,,,-.,..
> selector)
>
>
>
>
> wave
>
>
>
>
>
>
>
> i;)vJtll.~i:j.~ connectors
>
>
>
>
>
>
>-
>
>
>
>
>
>
>
>
>
>
>
>
>
EfO
costs are ... ·./Ui.CI'"".L~V'WI-
costs, W"IA . .el!1I· ..... 0~ .. i' or a year. 13
on
a
v,;:ll,..!UJUtiL,ll,lI;.V no """' .... """L .. Ji>., .....
........ ""l' ....... \'1·4?1l"l1l"'>' .. .:. cost n.a",~,,,'lF".""'<:'
a
6
cover installation,
to is as
>
>
>
> serves as an
>
>
>
assumes a rate
>
>
>
>
> a..~SUlnes a
>
>
>
> asslUlles n.o
:>
rate
Y"~'IC'''·'''''.n is $167
v ... """"" ........ at
parrs
at
>
>
>
> a
>
> assumes an on
> same as
>
> no
> can
> costs
> relnote
> consumes one watt a
2/Uleter 12/meter 1,15/meter
25
1 2
6
3 4
11
7
167
7 35
(79
16
30
3
9
7
~fibe:r
cable $hewitb
O!bk':~~tioo
h'm~rdud:
~Qr
LillMi. mwtf~ iawt
~f~~
WDM P~r
Con,~ ~meliltwiallt
T~.
mode f'ibet,
!lY\~th
C05ib~ iM.t.~tioo
l.noor duct
Comt~lt'
FIGUR.E63 INVESTMBNT COST
FTrH g N~ Ne~rlr
150,lOO 11,000
193,000 1,010
100
60,000
75
100
6Of 40 1,000
83,600 110
35
60,000
25
100
0.10,0.&5
1 1S 1
17
1 15
l65
1,498
140
1
1
1
3
1
Network
Component
Subscriber Premises
ePE (voice only)
Optical transmitter, optical receiver
ISDN chip
Digital - Analog converter
WDM
Cabinet
Hookup and installation
AC power outiet, battery backup
Total
Source: RAND Corp.
FIGURE 6.3 (Continued)
INVESTMENT COST
f<TfH - Narrowband Network
(In dollars)
CUlTcnt
Cost
60,40
15 7
35 10 SO/home
SO/watt
67
Future
Cost
167
SO/home
SO/watt
Current
1,182
50
150 1,382
3,900
Future
167
50
150
361
1.126
Network
Component
Optical source; optical receiver
Multiplexer
Power
Total
Feeder
Single mode fiber
Cable sheath
Cable installation
Inner duct
Splice, connector
Total
Optical source, optical receiver
Multiplexer
Line interface unit
Subscriber line card
Power
second drops)
C,ontroUed environment vault
Total
Distribution
Copper and cable
Cable installation
Inner duct
cable
Cable installation (buried)
connector
FlGURE6.4
lNVF..:srME"Nl COST
FITC - Loop Carrier
Current
Cost
750,300
11,000
IO/watt
l/meter
40,75
750,300
11,000
193,600
67
60,000
variable
50
0.50
(In dollars)
Future
Cost
60.40
1,000
2/meter
15,25
60,40
1,0(,'0
83,600
28 1S/watt
60,000
variable
50
0 .. 50
\...ost per Home Passed
Current
1
15 1
17
5
3
36
3
2
49
15 265
89 36
59
465
29 42
7
2
6
86
28
24 2
54
Future
1
1
1
3
2
3
36
3
45
1
114
37 36
59 248
29 42
7
2
6
86
28
24
2
Netwnrk
Component
FIGURE 6.4 (Continued)
INVE',,;;TMENT COST
FT1'C - Digital
(In doHars)
Current
Cost
F\lture
Cost
System
Current Future
--~---. ---.•.. ---.. --~--~--.-.-.---------.-------~.--.-.---.-. -------------
Protective block
Hookup and installation
'IOta}
36
96 132
36
%
ill
-------,--------_._-----------_. __ . -------,---~----. -------_._---
--._--------,--------------
Multiplexer
Power
Total
Feeder
Network
Single mode fiber
Cable sheath
Cable in.stallation
Inner duct
Splice, connector
Total
receiver
Optical transmitter, optical receiver
M uitiplexer
Line interface (framer)
Subscriber line card
Power - 8-hour backup supply
lbtal
Distribution
Single mode fiber
Cable sheath
Cable installation
Inner duct and connectors
Pedestal/manhole terminal
Total
Drop
Drop cable
Cable installation
Splice, connector
TOOll
FITC-
FIGURE 6.5
INVESTMENT COST
Narrmvband Network
(In dollars)
Current
('.-Ost
750,300
5,000
2/meter
12/meter
40, 75
750,300
5,000
20
50
O.20/meter
300
0.50
Ijmett:r
1.67/splice
75
Future
Cost
60,40
500
15,25
60,40
500
10
22
l/metcr
300
0.50
75
Co,;;t ner Home J:assed
Current
23
109 10
142
30
6
36 4
8
84
368
875
14
70 150
1,477
14
14
39
45
37
149
28
24
2
54
Future
2
11
10
23
15
6
36
4
3
64
84
88 7
31 150
311
7
14 39
20
37
n1
28
24
2
54
Network
FIGURE 6.5 (Continued)
INVESTMENT COST
l'"TfC - Fiber/Copper Narrowband Network
(In doBars)
Current
O)st
Future
Cost
Cost per Home Passed
Current Future
-----.---------~, ~~,,-----,-----------------.----------,~---
Protective block
Hookup and installation
Total
---------------------
Source: RAND Corp.
3D
80/home
30 30
80
110
2,017
30
80 110
619
-_._--,---
Headend Equiplnent Distributed video channels
transmitter Power
mode fiber sheath
Cable installation Inner duct Splice, connector Total
Optical Network Interface Optical receiver Amplifier, accessories housing Power 'row Distribution Cable plant Field electronics
components
O.lO/meter 2/meter 6/meter 1.5/meter
25
10/connector SO/splitter 4/meter 0.75/meter 1,000 3.2 km
10 3 1
1 6
3 28
1 2 5 8
36
19
68
l' ~,J
5 11
-------- -------- "'--
converter I-Iookup and installation Second converter
~--.--.----
100 SO/home 100
'"_.--------------
"" ....... "" .. Total
60 30 18
108
368
Cornponent
Video jukebox
Central Office Equipment Video switching, and control Broadband multiplexer, optical transmitter Voice multiplexer, trans., opt. Power Total
Feeder Single mode fiber Cable sheath Cable installation IImer duct Splice, connector Total
Remote Terminal (Serving,Area Interface) Broadband multiplexer, detector Remote video sWltching SLIC (4 subscriber/card) andWDM Voice multiplexer, opt. trans. receiver Power Controlled environment vault Total
Totals
Total distribution, loop, CPE Total electronics
nonelectronics
74
""""'.,. ... -......-.-------.-~
9 68 77
101 40
1 2
144
15 6
36 3 4
38 11
235
105 75 59
952 1,311
449
9
9
20 13
1 1
35
3 6
36 3 1
13 5
105 60 59
454
1,015 434
supplier to
Raynet has perforrned studies to
and architectures. A J"rvnnrr'l1
telephone lines.
such study ",.,.'-J A.U '-"i..!..'
.:u"-'" ... ,.....,..., was ...... HJu"""L'
were
bus
study with 160 active
of two fibers
were required. ilstar-to-the-curb" were required
fiber for each pedestal). the distribution portion network, costs and
prices appropriate to 1993 volumes were projected in 1989 dollars. Raynet estimated
the "star-to-the-curb" first cost for the portion the
network at $1,243 a line, and $885 the san1e portion of the bus
architecture. Raynet claims the lower cost bus system requires a smaller number of
fibers, consequently reducing the cost of splicing. Other fibers are not terminated
until required for additional services such as video; the noninvasive coupler
does a\vay with the need for splicing at the access point.1
New Construction
The majority of the subscriber loop " ...... .a .. ." ..... 'v .... u involve new
construction and the double star is the most
deployed. Most industry experts
only in new construction scenarios the 1993 to
architecture being
economical for POTS
frame. for
POTS plus video for existing networks will not be effective until much later,
estimation is based on the
video over a separate
LJAp;;".,.A.LA.'."" ..... ~.i., demand for other
sometime between 1995 2000, The reason
that it will be cheaper to provide POTS over fiber
frame.
TV must broadband services
network for both
1 Ibid.
"' ....... 1..L..J"-'."'" before a single fiber
'U'''-'', .. v<.J .... ll'\,.L and narrowband "'''''THIS"''-'''' viable.
75
7
AND OTHER PROCEEDINGS
The questions contained in the. three following sections are intended
function as guides the types of data a commission may need in assessing the
broadband deployment plans of jurisdictional utilities and are not necessarily formal
discovery questions.
Questions for Telephone Companies
1. Have you or do you plan to deploy coaxial cable or fiber-optic cable in
the loop plant for provision of broadband services? A) If so, how much of new
construction spending will be for fiber cable in feeder loop plant and how much for
subscriber distribution and drop portions of loop plant (annually for the next five
ten years)? B) Is the plan to use fiber-optic cable for feeder and some portion of
distribution cable (for example, between the SAl and RT or pedestal) and coaxial
cable for the last segment to the subscriber premises? C) If not, what is the
relative use of fiber-optic and coaxial cable that is contemplated? D) How much
would be spent on associated electronics for central office equipment, remote
terminals, pedestals, and subscriber premises equipment owned by the telco? E)
What new telco broadband service are contemplated? F) What is the time frame for
construction oisuch facilities? G) How many CUstomers per year do you plan to
provide broadband service to for the next five to ten years?
2. Is the primary motivation for your deployment of broadband loop plant
for provision of entertainment video services or other broadband services? What
are the other broadband services and what are your delnand estimates in terms of
new revenues per household· from broadband services five years into the future?
3. How much do you estimate that the perDsubscriber deployment costs of
broadband "capable" loops will be (regardless of whether they are actually used for
broadband services)? A) What do you forecast the trend in these costs will be
(separate by EF&I and OAM)? B) How do these costs with copper?
77
4. following provides demarcation points residential subscriber
loop plant between feeder (CO to SAl) and distribution (SAI/RT /pedestal) and
portions of subscriber loop plant Assuming a network configuration this type,
and assuming as well that a digital CO and digital SLC are in place, for the
following categories of loop plant, please classify those which are necessary
provision of broadband services (such as cable television), and not for other
narrowband services (that is, traditional or new narrowband services): A) For
equipment and devices located at CO, RT, pedestal and subscriber premises: WDM,
channel selectors, lasers, OlE and E, i 0 conversion devices, optical bridgers, optical
connectors, optical mix, channel banks, signal transmitters, receivers, detectors,
codees, line cards, and so on. For transmission facilities in feeder, distribution
(shared vs. dedicated) and drop: coaxial cable and fiber-optic cable.
5. Would there ever be a need for using coaxial cable in subscriber loop
plant unless broadband services to residential customers are contemplated? Is
coaxial cable deployment contemplated for anything other than video services?
6. Under what conditions would you deploy fiber-optic cable in dedicated
portions of subscriber loop plant? A) Why? B) At what point in the future would
fiber optics for dedicated subscriber loop plant prove to be cost justified over
copper? 1) For new construction? 2) For rehabilitation? 3) For growth? C) What
is the contemplated useful seIVice life of broadband loop plant for feeder and
distribution portions? D) How does this compare to new narrowband facilities lives
(book and tax lives)?
7. How do you classify fiber-optic and coaxial cable costs when they are
used in subscriber loop plant? A) What would it take to separate them into shared
feeder, shared/dedicated distribution, and drop categories for tracking purposes? B)
How about electronics and equipment and devices used for broadband services like
cable television? C) Do you plan to alter your system of bookkeeping to be able to
track such costs between those required for narrowband services and those required
for broadband service capability?
8. What are the per-subscriber costs for each residential broadband service
trial currently underway? A) What are such costs forecasted to be as deployment
continues? B) What demand and revenue forecasts have you made for broadband
services? C) What is the source of funds for construction and how are they
accounted for?
78
1. already done so? yes, in what part of your
or network is fiber-optic cable
being installed or planned? Why?
2. your system fully addressable so that may identify bill specific
households for premium services? Is it two way-addressable so that
customers may enjoy payaper-view- type services? B) not, when you eXlJect
to upgrade to two-way addressability? 3. Do you plan further network upgrades for some type of two-way signaling
and services(s) such as videotex (for example, shopping-at-home), distance learning,
data, audiotex or voice? When do you expect to be able to perform such upgrades
to offer such services? (Please be as specific as possible)
4. Are you involved in any field trials of new two-way services? A) Do you
have any planned? B) Hyes, is the local telephone company involved also? C)
How are you interconnected with one another? D) Do you have any network
schematic diagrams or illustrations of two-way service configurations? E) What type of equipment is required for two-way services on your cable system? F) How
much does it cost per subscriber? G) What are your estimates of such costs (or
even the trend) over the next five years?
5. Do you believe cable systems will ever evolve to the point of providing
some types of two-way services for telecommunications, beyond that which will
likely occur for pay-per-view video? A) Do you believe that on-demand video (for
example, electronic video Hbrary instead of the local video tape store) is a viable
service offering? B) If yes, when do you think it would be available to your
customers? C) Is there a big future market for these types of services? D) How
about two-way voice and data services?
Questions for Equipment Vendor:i
1. What systems and equipment do you have available for purchase by
telephone companies provide broadband telecomnlllllcations services, including
traditional cable television-type service? A) Are your systems/equipment complete
or must they be used in conjunction with another vendor? B) If so, what part do
79
to
equipment
future n·rl .... 'O'·!i:'
systems/equipment over the next
acceptable )?
5.
or
expect to
6.
these
is the
current or estimated
'-'v ........... v' ...... and esthnates are
lllany orders
you
telephone companies to
'"''-'"JU ... ''-' ... ''3 ..... .<...,''' to two-
for business or
still
degree as
to saIne
¥ll"1:li~t~ P'!1f' }\OWl\$'I'I(ll.d ~!bl!rj'" Cost'l\, ~iS;'(",\\"\L'I, 19157'·,000
(C~Ui1t Ool
CMti"'!Ii!. Off i e~
$\!ii ten
1i<l!!OOl;l1l fi!lMllil'IkBUC!V
:)~·l'1li'ClIiIiv.,r
OLe ~~. Tr~c~iv~r3
Con'r$ '/01" YOM$ EnclOIMJrlt
DistJ"ib.rtion
F i blitr
;!~'i!!!!!'eat
CIrCle
'foul C':)lH~
Total ~OOQ Costs
199'5
$1,7['0
S1,~
Prej~~~ Fiber Ca~t3, POiS'~\U$,vld~, 1987·2000
(C~t~ne OQlt~r~J
CM'!ti"lIl OH iC:1t
$wi tl:;;f1
'ilfiffll:ttfi hf'l'lllNl1UaY
O~·;tJ!I;Ctli"'~r
CrO$$'e~~
Vl~ O;":t. Swi ten
~~~. irana~~jv~i"$
eM; f'S &/or Y:JI>l!ll
Erclowrlll
()1 lltril::~.rti on )I1~l'"
jl~u!'
Orem
!<:tul. COIlU
\"000 CQ.:I!.U
$ns ~1S0
SS25
$1,000
~O
~O
12,100
$1,rM
$7,~05
8
:'1.:2,670
~.,J70
2000
2000
$10
$55
$.2 .. OSO
Sl,e10
~f'oj~t~ Fi~r Sa(;i(~ Cost!!,
Ov~r\'<llY C~st$
"irart$.l'l'lii tti!11'"
F lo.r
tMt:H lc@t!on
Recllii"'~f'
ioUi. Ov\lJit!"t:II'(
tmullation ~~it e!"':!
C.Ollilts
'foul, SJlIvinqs lne~~tal i~f' C03tS
1988
119.33
S1.n
SJ.08
~"41..
130.57
ilCt!l>!i Pt'lI3Sed,
i'XiO
::;'5.
S'Ln
:tJ.08
.22 s1.31
Cabll!!! Telev"isicn Flbet· ':0 ;:hl!l KO'I'Iat, Cost: P-er
(Constant Oottars)
riber a~cxbone T ran$m1 e--::!ill"
fbltf'
(i'l31!a llat:; 0fI
~K.1Vl!»f"
rOU\ aK~
Distf"iwtion if1f'1smittftf'
Fl~f"
KJte.i".r SUr' Cougtlif'
TQUlO htl"ii::a.nd on
65~
$163.64
i:305 $763.64 .$J!~ is
~1,a1O.9t
:$10(LOO
,m.os
84
1968-2.000
i995
'S:Z~5a
$1.72
'U.oa $1.53
$3.92
2000
$1.93
S1.n
13.03 $1,.2.9
$8.02
~5.U
1~.80
~f", 1990-2000
2000
SJ.7Q Sl.~
:t2;47 $2.21
$4,t..2 $3.94
$:2.22 $1.65 $12.81 110.26
~.59 $19'2.25
$23 1.21 S192.25
$466.53 3256.34 Su,.n $19,2,3
,193.25 ~.O7
$10CLOO S100.00
,3Q6"Q,6 $rnl.3S
Cacle rel~~i ion Fi~r to ,n~ ~~,
(CQn$t~~t Oo~iar~)
19&3 19',10
Cable P~~tration 65'X
Fiber B,3t':~
ir<ll!"l$l'llitt!t" $5.15 F ibttr $1.72
instaLlHion S3.oa ;:{l!teeiv'I!Jf' 12.58
Toul a.u;;x~ 112.53
J1sttlbJt:ion
Transmit'!:!!" $500.0,0
Fiber $200.00
~eceiv~l" :5500.00
Sur C~lef" ~.
iout 01stribution $1,7.1S.00
Ca.IlIlt 01"00 $100.00
iotal C,:J:St 51,Yo57.53
85
199'5 2000
'fin:; 7S%
~.5a $1,93
S1.n Sl.n
s3~Oa $3.07 $1.55 $1.29 s.a.n $8.02
$312.50 S150.00 $.175.00 S1S0.CQ s3Z:S .00 $<'00.00
S1S.7'S $15.00 1-831.25 S515.00
1100.00 S100.00
S'94C.17 $.623.::l2
To investigate the potential evolutionary terns for combined t ansport systems o a s Ie cost projection model wa,"~ develc us of experience curves.
that over t costs decline
bed as vo it is empi
and
Using experience curves r 1 requires qualifiers. Fir t when app1i ca does not 81 for rovements ionali is of part ar importance in the case f fiber ics*
have many t s the functionali 0 past terns, as as ower costs. Also, the effect on costs is caused not increased volume, but tment in technol which leads to lower costs. c asses have ~ tten 0 m
learning curve fail to up n manufactur ng and product technology. Fina ly, since the Ining curve relies on accumulated experience, the of 1 application declines when the rate of accumulation dec il1e5.
In this tigation, of several kinds of resi ial fiber ic systems have been priced out based on projected price levels for 1988 at hi levels of aggregation. Major sections of the ana is were:
needed to support the tems l exclusive of local access
other common items. Va f Ii
86
and site costs red" Site costs
on ,000 per
needed for system: cou.p I etc#
aced spl , spl
onooxes, etc~
Hardware costs associated with the home Includes drop cable. ellS r
in-heme distribu ien cable. etc
A series of assumptions was made eg rd n9 projected ration 0 these ear hi t systems Re og on
predictive appl tieD e experience cu hniques and on the assumed volume projected costs were es abli hed for units at the end of the forecast periods.
In order to investigate the i luence of changing assumptions three variations of the basic model were completed. We rounded to the nearest full percent.
In first case, the Dfluence of ear techno was simul ted increasing 10\ to 2 of the new construction over This had the result of lowering the ear
ance of the penetration from
per od 1990-1995. em casts by 13\.
Next the act roved later acceptance was evaluated ch ng the later ration from 2\ to 5\. This had no on ear costs but resul n an rovement to the base case of 14\ in later casts.
Next. the potent a1 t of roved technology as reflected in a better experience curve rate was simula ed lowering the rate about 6\. This esulted in an rovement in the early costs of 9\ relative to the base case and of 14\ n the later cost.
Final I we did a simulation in whi all three factors were assumed: increased ear penetration, increa ed late ration, and oved technology, Th s produced uns antial gains: a 24% reduction in average per-subsc iber cos s n 1995 to $1,700, and a 23\ drop in 2000 to $1000 c~mpared to the base case. Cost r of this magnitude illus rate the potential beneficial
a regulatory policy framework that is especial in allowing the pro life ation of combined transport
8
Base case Ie star J' 00 , 0 $1,,200
5i Ie star 5,100 .2,500 1#300 Bus
r 0 ,300 $1,400
tration star $400 $1,900 $1,200
5,,100 20100 1,300
Aver ,600 ,000 $1,400
tet penetration Double star ",400 $2,100 $1~200 Single star 5 p 100 2,50'0 1#100 Bus 2,2QQ
:r $4 600 $2,300 $1, 0
st reduction rate Double star ,400 $1,900 $1, 0 Sing Ie star 5,100 2 / 200 1,100 Bus 2/1QQ _ 1 . 30Q
Aver e .. 600 $2;100 $1,200
Combined f Double star $1,600 Sing star 1,800 Bus 1 c6.QJl
1'a I 00 $1 700 .. 000
( to nearest $1(0)
88
60 117 246
30 47 470
10 6 21
4 2 1
33 SO
Mult:i:oiaa 117 70 LED. PIN RCVR WDM 105 -VideoCoder,~ .... ~ 138 83 Rcinforced T erm.i.nal 7 4-
Tota) 367 220
219 131 1139
125 125 H ()()l( -lID and I.n.s:t3..llari 50 50
Total Overall Total 88
89
11
310 262
g 35 4
20. 27..5 10
Totlll 32
191 3
122 7
313
1558
12.5 125 75 50 50
Tota f 115 lOS Total
Backbone ::::
90
w
o
91
. j
j J
j
j j
j j j
j j j j j j j j
j j j j j j j j j j j j j j j j
j j j j j j j j j j j j j j j j
j j j j j j j j
j j j j
j j
j
j
j j j
"Jones System Its 1989.
Projects.'l Fiber
12,1987.
Bell
Loop." Proceedings) Fiber 1990.
Say is Only 1990.
III, "A High Video 1980.
Top Related