Government Policy vs. the Fiber-to-the-Home Supply Chain
Andjelka Kelic
Massachusetts Institute of Technology
N42-140U, 77 Massachusetts Ave. Cambridge, MA 02139
617-253-1436 Fax: 617-258-9535
sly@ mit.edu
Abstract
A policy for rapid deployment of fiber-to-the-home may be in direct conflict with the health of the
transceiver component supplier industry. The interests of consumers, regulators, and even
service providers are in conflict with the industry that provides a critical component necessary
for the service. The industry needs to recognize this conflict and explore strategies to keep itself
viable in light of these conflicts. A system dynamics model is used to explore the effects of
government policy on the deployment of fiber-to-the-home as a broadband technology.
Specifically this article investigates the effects of a policy for rapid broadband deployment on the
component supplier that is farthest from the consumer in the supply chain.
Introduction
Fiber-to-the-home (FTTH) refers to the provisioning of narrowband and broadband services to
residential customers over an optical cable rather than traditional copper wiring. Traditional
telecommunications companies, such as the Regional Bell Operating Companies (RBOCs) of
Verizon, Qwest, SBC, and BellSouth have an interest in fiber-to-the-home technology to help
them compete with cable companies for the delivery of voice, video, and data service. Fiber-to-
the-home technologies are especially of interest in areas that do not yet have broadband service
and are out of reach of DSL (Digital Subscriber Line) service.
On March 26, 2004, President George W. Bush called for universal affordable access to
broadband technology by 2007 (Keto, 2004). The system dynamics model presented here
explores the effects of that mandate on the industry that supplies transceiver components for
fiber-to-the-home. For the purposes of the model, the mandate means that all of the communities
that do not yet have broadband will deploy it within three years. To explore the potential best
case for high volume of transceiver components, the model assumes that all future broadband
deployments are of fiber-to-the-home.
Technology Overview
An understanding of the technologies used to provide fiber-to-the-home service helps to illustrate
where the transceiver suppliers fit into the picture. Fiber-to-the-home technologies fall into two
categories: active and passive optical networks. Active optical networks have an active
component (such as a switch or a router) between the central office and the end-user. These are
point-to-point networks with switched traffic, as shown in Figure 1. This network is similar in
architecture to traditional hubs and switches that run local area networks.
In an active network, fiber-to-the-home transceivers are located in the device at the customer
premise (customer premise equipment or CPE), and in the customer facing side of the active
switch. Thus for every additional active optical network customer, two transceivers are
deployed.
Feeder link Distribution links sees
| OL | Customer 1
nooo
Acti itch Customer 2 |
z ctive swil ¥
Central office (Remote Node) =>
Customer N
Figure 1; Active Optical Network
Passive optical networks (PONS), or passive star topologies, as shown in Figure 2, have no active
components between the provider's central office and the subscriber. PONs are point-to-
multipoint systems with all downstream traffic broadcast to all customers. The majority of fiber-
to-the-home technologies being developed and deployed today are passive. In these networks,
the fiber-to-the-home transceiver is located at the customer premise.
Currently, the transceiver devices being deployed in passive optical networks are different than
those being deployed in active optical networks. In passive networks, there is also a high
powered transceiver located in the central office. This device is sufficiently different in
performance requirements from both the active transceiver and the passive customer premise
transceiver that it is not included in the transceiver deployment rates discussed in the model. For
a new customer on a passive optical network, one additional transceiver is deployed.
Feeder link Distribution links f=]
= | = Customer 1
: 0 —
Customer 2
Central offi y
emia omce Optical Splitter ea M
(Remote Node)
Customer N
Figure 2: Passive Optical Network
High-Level Dynamics of F iber-to-the-Home Deployment
Figure 3 shows a high-level overview of the dynamics governing fiber-to-the-home deployment.
The text blocks represent variables associated with broadband and fiber-to-the-home adoption
and deployment. The arrows between the variables represent causal links. As a variable
increases or decreases, the variable that it is linked to changes in the same (+) or opposite (-)
direction. For example, an increase in “available content and applications” results in an increase
in “broadband customer base.” Alternately, a decrease in “available content and applications”
would result in a decrease in “broadband customer base.” The green arrows represent dynamics
that exist in the world today.
Currently, residential broadband subscribership is growing. Cable modem and DSL service are
reaching more and more communities. The green loop entitled “content brings users” represents
customers bringing more content to the network in the form of user-provided content and
provider-based content (both commercial and non-commercial). Content in tum attracts more
user interest.
/ network extemalities (community)
4 =
\ fiber-to-the-home ~ SS
| netvod enteriies (commercial) demand 42s) ~
)
) demand drives community deployment
community
broadband £4) fber-to-the home
istomer base leployment
ena demand drives commercial deploym :
ts) +commercial
yl = deployment
content brings users 2
Ss :
fiber to-the-home 4 cost to provide
customer base Cy fiber-to-the-home
ey cost drives commercial deployment Senice,
available
content and
fiber-to-the-home + te
applications iber-to-the-home + fiber-to-the-home
revenue > technology innovation
and cost reduction
Se)
cost drives community deployment.
Figure 3: High level drivers for fiber-to-the-home and broadband deployment. Green
represents effects that are active in the market today; orange the dynamics that are just
beginning to emerge; and red represents hoped for dynamics.
However, not all areas are considered economically viable for broadband deployment. In many
areas, the cable infrastructure is older and incapable of supporting cable modem service without
complete replacement. Other communities are sufficiently spread out that DSL service cannot
reach all residents, or the presence of loading coils and degrading copper wires makes service to
those areas not viable.
The communities that do not have service are represented in the green loop labeled “demand
drives community deployment.” Communities without broadband service feel more pressure to
acquire it as the number of communities with broadband increase. Those with municipally
owned utilities or served by rural local exchange carriers that have been left behind by the latest
wave of broadband deployment are beginning to look towards fiber-to-the-home as a solution for
residential broadband service.
Fiber-to-the-home, like other telecommunications technologies, is subject to network effects.
The more people that have fiber-to-the-home, the more interesting it is to others that do not yet
have it. The same effect applies to broadband service overall, just as it does to narrowband
services like telephones. This effect, shown in the orange loop labeled “network externalities
(community),” is just starting to emerge in the market as broadband growth becomes more
visible and communities perceive it as a key to economic success.
The other loop that is just beginning to emerge, and is of particular relevance to the transceiver
market, is the orange loop labeled “cost drives community deployment.” Cost reduction in the
transceiver and electronics equipment market is making fiber-to-the-home more attainable to
community deployments. These deployments in turn are producing additional volume for the
industry and economies of scale which drives down cost.
It is hoped that the emerging orange cost reduction loop from municipal and rural deployments
and the emerging demand loops will activate the currently dormant red loops of “cost drives
commercial deployment” and “demand drives commercial deployment.”
Overlying this already complex set of interconnections is government regulation at the federal,
state, and local level. The interconnections between all the pieces of the broadband deployment
puzzle are so complex that effects of policies are unknown before implementation and/or have
unanticipated side effects. For example, TELRIC pricing, the FCC method for calculating
wholesale rates for network elements, was intended to encourage competition. However, it
appears to have discouraged infrastructure investment by both incumbents and new entrants.
The three year mandate expressed by President Bush included no details about how the policy
would be achieved, so the model simply assumes that deployment happens to all communities
that do not currently have broadband in the three years following the issuance of the mandate.
In a real world scenario this would be analogous to the government forcing the RBOCs to deploy
broadband through some sort of incentive similar to the Universal Service Fund for telephony.
Since the RBOCs are more heavily regulated than the cable companies, they are the logical
choice for a government mandate. In most cases, the areas that do not already have broadband
service are out of reach of traditional DSL, thus fiber-to-the-home would be the technology of
choice. Assuming that all new deployments are of fiber-to-the-home serves as a somewhat
extreme case test of the effects rapid universal access would have on the fiber-to-the-home
supply chain.
Implementing the Mandate
The mandate is implemented in the model through the stock and flow mechanism shown in
Figure 4. The model takes the number of potential communities and the mandated completion
time and calculates how many communities need to deploy in a given timeframe to meet the
mandate. The stock of “mandate deployed communities” is used to calculate broadband
availability to households based on an average number of households in any given community.
mandated
completion time
necessary
deployment rate
mandate initial mandate
potential communities deployment time mandate initial FTTH
aTETy communities
potential = mandate
communities : P| deployed
mandate deploying communities
communities
Figure 4: Mandate build stock and flow diagram.
The resulting broadband availability in the default run of the model is shown in Figure 5. The
term “other” in the legend of the figure refers to pre-existing cable modem and DSL service in
communities. Since the model assumes that all new builds are of fiber-to-the-home service, the
availability of DSL and cable modem service to households does not change from its initial
value. The initial values for the model come from summer 2004 data on communities that have
cable modem (Warren Communications News Inc., 2004), DSL (broadbandreports.com, 2004;
North American Numbering Plan Association, 2004), and fiber-to-the-home service, (Fiber-to-
the-Home Council & Telecommunications Industry Association, 2004) and the number of
subscribers to the services (Federal Communications Commission, 2004).
Figure 6 shows the causal diagram along with stocks and flows that converts broadband
availability and potential customers to actual customers of broadband service. Households with
broadband service available to them are placed in a stock of potential customers based on their
willingness to pay the current price being charged for the service. Prices sensitivity was estimate
based on the current average price of cable modem service in the Unites States (Warren
Communications News Inc., 2004) and consumer price sensitivity information (Ainscough,
2003).
In order for households to actually adopt the service that is available, they need to be made aware
of it. Two awareness mechanisms exist in the model: advertising, and word of mouth. The
mechanism shown in the figure is adoption due to word of mouth. Customers that already have
service interact with those that do not, and at some rate, convince the potential customers to
acquire service. The conversion rates used in the model were calibrated from residential
broadband data published by the Federal Communications Commission (Federal
Communications Commission, 2004). The model assumes that new customers are more excited
about their broadband service than customers that have had the service for a while, or customers
that are returning to the service after having left it. So a new customer is more likely to attract an
additional subscriber to the service then a customer that has had the service for a while.
80M
60M
40M
20M
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
calculated households with broadband available[other] : mandate households
calculated households with broadband available[FTTH] : mandate ———— households
Figure 5: Community deployment rate under a three year mandate.
oldwom 4 Tew WOM!
Conversions ei Conversions
total wom
conversions
‘ new assimilated
Poe buyers os customers
TAY Onleity new product assimilation rate —
amtival initial new buyers 4 aggiltttea
time to adopt time to assimilate
static delivery
delay
Figure 6: Customer broadband adoption.
Figure 7 shows the characteristic trend of the user adoption market for broadband. Initially,
adoption is slow. Few people have broadband service, and even fewer have fiber-to-the-home
service, so there is not a strong network effect, and marketing drives user adoption. The model
assumes that the network effect that governs adoption is for all broadband, thus if a fiber-to-the-
home customer interacts with a person that only has some other form of broadband available,
that person is likely to adopt the broadband service available to them. As can be seen in the
graph, once a critical number of users have signed up for service, adoption grows rapidly. The
rate of adoption begins to slow once all communities have service available, and then reaches a
limit.
40M
30M
20M
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
broadband households[other] : mandate households
broadband households[FTTH] : mandate —-------———_ households
Figure 7: Consumer adoption trend for a three year deployment mandate.
Implications for the Transceiver Industry
Background research relating to the deployment of fiber-to-the-home and the state of the
transceiver industry showed the following (Kelic, 2003):
e A large proliferation of standards for fiber-to-the-home transceivers;
e No clear convergence path or natural advantage for any one fiber-to-the-home
technology;
e It is impossible to predict a ‘winning’ fiber-to-the-home technology and the direction of
standards evolution; and
e It may be possible to standardize on one or two transceivers for both customer premise
and remote terminal equipment.
These factors are translated into the following assumptions that are used in the model to go from
customers as discussed in the prior section to the transceiver deployment rate:
e fiber-to-the-home deployments are 50% active and 50% passive;
e active deployments use the same type of transceiver at the remote terminal as at the
customer premise;
there is a five year equipment renewal rate;
only one CPE is deployed per customer;
e the equipment being deployed has greater capacity than required for customers, so there
is no driver for equipment replacement aside from equipment failure; and
e Standardized transceivers can be used at the customer premise for passive deployments
and at both the customer premise and the remote terminal in active deployments.
Figure 8 and Figure 9 show how transceiver tracking is implemented in the model. The
deployment rate of standardized transceivers is a simple sum of the transceivers in the remote
terminals and those in the customer premise equipment. The deployment rate for active and
passive transceivers is calculated using the percent active and percent passive deployment rates
specified above.
<CPE deployment
rate>
see
standardized xcvr Bb — failure rate
deployment rate pea XCV :
4 deployment>
<remote terminal port <remote terminal port
deployment rate> <CPE failure failure rate>
rate>
Figure 8: Standardized transceiver deployment rate.
<remote terminal port <CPE deployment
<standardized xcvr deployment rate> rate>
deployment rate>
<% PON>
xcvr deployment.
me + — & active>
Figure 9: Transceiver deployment rate, all transceiver types.
The transceiver deployment graph that results from the market assumptions and the transceiver
assumptions is shown in Figure 10. This graph shows that transceiver volume grows as
deployment is happening, but then drops off to a replacement rate for equipment that fails. This
effect is due to fiber-to-the-home deployments not requiring the leading edge in transceiver
technology. Users are not demanding bandwidth anywhere near the capacity of the transceivers.
This gives the carriers the ability to upgrade bandwidth by making changes in software without
needing to replace equipment. The capacity of the equipment is much higher than current
demand, and even near-future foreseen demand.
As shown in the figure, standardizing transceivers for fiber-to-the-home deployment sees a faster
growth rate and a higher peak, and a similar decline to replacement rate. Standardizing a
transceiver just for the fiber-to-the-home market is not enough to prevent the growth and decline.
This sort of growth and decline can be devastating to an industry, since it requires the industry to
build up capacity to meet demand, however that demand is not sustained. The excess capacity is
likely to cause financial difficulties in the industry and individual companies to fail.
Since the rapid growth and decline comes about due to a fictitious policy that assumes that
telecommunications carriers will be able to rapidly deploy fiber-to-the-home, it is necessary to
explore the effects of a slower deployment rate on the transceiver industry. A slower
deployment rate, while not in the consumer or regulator’ s interests, may help mitigate the growth
and decline and also better reflect the constraints of an actual infrastructure build.
400,000
300,000
200,000
100,000
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
xcvr deployment rate[ standardized] : mandate ———————————_ uniits/Month
xcvr deployment rate[active] : mandate units/Month
xcvr deployment rate[passive] : mandate —————————____—_ units/Month
Figure 10: Deployment rates for standardized, active, and passive transceivers under a
three year mandated deployment schedule.
Slowed F iber-to-the-H ome Deployment
If the deployment of fiber-to-the-home is slowed to a ten year deployment rate, still requiring
that all of the communities that do not currently have broadband build fiber-to-the-home, the
customer growth rate looks like that shown in Figure 11. As shown in the graph, it takes
significantly longer for consumers in areas that do not already have broadband to adopt
broadband technology. Delaying the deployment of broadband also causes adoption by people in
the areas that already have it to slow slightly, since fewer customers causes the network effect to
not be as strong.
The slower infrastructure build rate and adoption rate results in the transceiver deployment rate
shown in Figure 12.
40M
30M
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
broadband households[other] : slowdep households
broadband households[other] : mandate households
broadband households[FTTH] : slowdep households
broadband households[FTTH] : mandate households
Figure 11: Customer Growth Rate Under a Three Year Versus Ten Year Deployment
Schedule
400,000
300,000
200,000
100,000
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
xevr deployment rate{standardized] : slowdep units/Month
xcvr deployment rate{standarized] : mandate units/Month
xcevr deployment ratefactive] : slowdep units/Month
xevr deployment ratefactive] : mandate units/Month
xcvr deployment rate{passive] : slowdep units/Month
xevr deployment rate{passive] : mandate units/Month
Figure 12: Transceiver Deployment Rate Under Ten Y ear Fiber-to-the-Home Deployment
Schedule
As shown in the figure, slowing the deployment rate does make the peak less dramatic and also
lengthens the build-up period before the peak. Slowing the deployment rate even further, to a
twenty year infrastructure build time, results in the user adoption rate shown in Figure 13. The
corresponding transceiver deployment graph is shown in Figure 14. Delaying the deployment
completion time even further from ten years to twenty years drastically reduces the peak seen
under the three year mandated deployment schedule.
40M
30M
20M
10M
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
broadband households[other] : slowdep20 households
broadband households[other] : mandate households
broadband households[FTTH] : slowdep20 households
broadband households[FTTH] : mandate households
Figure 13: User Adoption Under a Twenty Y ear Deployment Timeframe.
400,000
300,000
200,000
100,000
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
xevr deployment rate[standardized] : slowdep20 units/Month
xcevr deployment rate{standardized] : mandate units/Month
xevr deployment ratefactive] : slowdep20 units/Month
xevr deployment ratefactive] : mandate units/Month
xcvr deployment rate{passive] : slowdep20 units/Month
xcvr deployment rate{passive] : mandate units/Month
Figure 14: Transceiver Deployment Rate Under Twenty Year Fiber-to-the-Home
Deployment Schedule
A slower deployment rate may be better for industry health in the short term because it delays
the drop to a renewal rate and reduces that drop. However, it also results in much smaller overall
production volumes for the industry and is far worse from a consumer perspective. Consumers
have to wait much longer to be able to obtain broadband service, which is not in the interest of
the consumer or of the policy makers.
Disposable Equipment and High Churn
Another potential way to prevent the ramp up and decline in transceiver deployment rate is a
high customer turnover rate along with disposable customer premise equipment. If customer
premise equipment cannot be redeployed by the service provider after a customer cancels
service, that piece of equipment is disposed of. Any new customer additions or re-additions of
former customers then require a new piece of customer premise equipment and a new
transceiver.
Figure 15 illustrates how the discarding or redeployment of equipment is handled in the model.
In the default runs of the model, the discard switch shown in the figure allows CPEs to be
redeployed from customers that have left the service. When the switch is tumed on, CPEs from
customers canceling their service are discarded and not redeployed.
replacement C PE
total deployed ?
CPE deployment CPE CPE failure rate
rate e. ,
new CPE ‘ time to fail
deployment PE removal due to
ee CPE cancellation
redeployment rate
wv
time to redeploy ‘4 ss vel
discard switch CPE
Figure 15: C PE redeployment and discard.
The results of disposable equipment under the original set of conditions for a three year
mandated deployment schedule are shown in Figure 16. This deployment assumes a thirty-five
percent churn rate (Credit Suisse First Boston Equity Research, 2003). The churn rate is the
fraction of customers that leave the service provider annually. The default assumption is that
fiber-to-the-home will have a chum rate similar to that of cable modem service since the
characteristics of the installation are similar.
As seen in the chart, forcing service providers to discard customer equipment with a thirty-five
percent customer turnover rate is extremely beneficial to the transceiver industry. The peak still
happens, but since the customer tum over rate is fairly high, much of the deployed equipment
gets discarded before it fails, so new equipment deployment overwhelms the replacement rate.
600,000
450,000
300,000
150,000
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
xcvr deployment rate[standardized] : discard ———————————__ uniits/Month
xcvr deployment rate[active] : discard units/Month
xcvr deployment rate[passive] : discard units/Month
Figure 16: Transceiver Deployment Rate Assuming Disposable Customer Premise
Equipment.
A high customer chum rate is costly to the service provider since there is a cost associated with
acquiring each new customer. Service providers are investing significant effort into improving
customer satisfaction to reduce the churn rate. The effect of a chum rate lowered to ten percent
is shown in Figure 17. As can be seen in the chart, the interest of the service providers in
lowering chum is in direct conflict with the interest of the transceiver industry in eliminating the
growth and decline. Even with disposable equipment, the lower churn rate offsets the benefit
gained by requiring that any returning customer get a new piece of customer premise equipment.
600,000
450,000
300,000
150,000
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Years
xevr deployment rate[standardized] : lowchum+discard units/Month
xevr deployment rate[standardized] : discard units/Month
xcevr deployment ratefactive] ; lowchum+discard units/Month
xevr deployment ratefactive] : discard units/Month
xevr deployment rate{passive] : lowchum+discard units/Month
xcevr deployment ratelpassive] : discard units/Month
Figure 17: Transceiver Deployment Rate Assuming Disposable Customer Premise
Equipment and a Ten Percent Churn Rate.
Summary
Government can have a devastating effect on the health of the transceiver industry. A
government push for rapid deployment of fiber-to-the-home without an industry move towards
standardization across product lines (telecommunications and storage, for example) could be
devastating to industry health. This sort of push causes a rapid boom and bust cycle, with
industry rapidly increasing capacity to meet demand and then left with a state of overcapacity
after the deployments are finished.
Moderation in the rate of deployment would make the boom and bust cycle for fiber-to-the-home
transceivers much less severe. However, slowed deployment is difficult for the transceiver
industry to control. Increasing transceiver prices may slow fiber-to-the-home deployment;
however it may also force service providers to a different technology, resulting in lower
transceiver volume. Slowed deployment is not in the interest of consumers or regulators, since it
delays the availability of broadband technology to the consumer.
Regulators at the federal, state, and local level have an interest in promoting telecommunications
technology and ensuring that the technology becomes available to all consumers. When these
regulators set policy they examine the problem from the perspective of the consumer and
consumer choice. This results in policies that are designed to benefit consumers from a price and
service perspective, and also to assist new entrants in providing competition in the market.
Rarely do telecommunications policies look beyond the consumer and the telecommunications
companies to the effects on the remaining portions of the supply chain. The transceiver industry
is at the opposite end of the supply chain from the consumer and thus subject to consequences of
telecommunications regulation. Recent broadband policies have been designed to promote the
rapid expansion of broadband services to the consumer and also to promote facilities-based
competition. However, rapid deployment of broadband translates down the supply chain to a
need for large production capacity that gets used in the initial network build, and then sits mostly
idle used primarily to replace equipment that has failed.
The regulatory viewpoint of watching out for the good of the consumer is unlikely to change.
The industry needs to explore ways to protect itself from the cyclical nature of the
telecommunications industry and prevent situations of overcapacity and excess inventory.
The state of the transceiver technology makes it difficult for the industry to avoid a boom and
bust cycle. Fiber-to-the-home deployments do not require the cutting edge in technology,
especially at the customer premise. All of the existing standards deliver far more bandwidth than
the consumer can currently use. The systems are designed so that providers can implement
upgrades in bandwidth through software, without having to swap out components. For
transceiver manufacturers this means that once initial deployment is passed the replacement
driver will be equipment failure and replacement as opposed to technology upgrade, leaving
excess capacity from the deployment ramp up.
In current broadband deployments, customer premise equipment has an expected life of five to
seven years. This time frame is far shorter than the twenty-year standard for the industry’s
traditional telecommunications product line (transceivers in the backbone of the Internet), yet it
is still too long to prevent overcapacity once networks are deployed. Service providers have
reduced their reliability expectations for equipment deployed at the customer premise to be more
in line with that of consumer electronics than with traditional telecommunications networking
equipment.
Customer premise equipment that could not be redeployed in the event a customer left the
service coupled with a high tumover rate of customers would help mitigate the fall to
replacement shown in many of the transceiver deployment graphs. However, this could
potentially be devastating for the telecommunications provider, since a high turnover rate means
less cost recovery of the expenditure to acquire the customer and lower revenue. Equipment that
could not be redeployed would also increase the cost to acquire any individual customer and the
cost of the service as a whole. An option of equipment that is disposable is not entirely in the
control of the transceiver manufacturer, since in many cases they do not manufacture the
customer premise equipment itself, just the transceiver used in that equipment. Therefore such a
transceiver industry policy would require the buy-in of all equipment manufacturers and
telecommunications providers to be viable, and there is no incentive for cooperation on this
particular issue. Customer turnover rates are also not in direct control of the transceiver industry;
they depend on the ability of the provider to attract and retain customers.
Facilities-based competition with multiple providers building fiber-to-the-home networks would
perhaps mitigate the boom and bust cycle. However, the networks are costly, and to date the
majority of facilities-based competition has not been occurring in identical technologies. For
example, broadband is currently provided over traditional phone networks (DSL), cable
networks (cable modem service), and in some cases wireless networks.
Standardization across markets is one of the potential solutions that is under control of the
transceiver industry and also good for industry health. A standard transceiver that can be used
across markets would protect the industry from the cycles associated with telecommunications
deployments and opens up the potential of appealing to additional markets. The transceiver
industry needs to aggressively explore standardization across product lines to bring itself in line
with the interests of the rest of the fiber-to-the-home supply chain and ensure its survival
regardless of the direction that fiber-to-the-home deployment takes.
Bibliography
Ainscough, D. (2003). The realities of tiered broadband service. Boston: The Y ankee Group.
broadbandreports.com. (2004). Central office finder. Retrieved August 19, 2004, from
http://www.dslreports.com/coinfo
Credit Suisse First Boston Equity Research. (2003). The broadband battle: Credit Suisse First
Boston.
Federal Communications Commission. (2004). High-speed services for Internet access: Status as
of June 30, 2004. Retrieved December 26, 2004, from
http://www.fcc.gov/Bureaus/Common_Carrier/Reports/FCC-
State_Link/IA D/hspd1204.pdf
Fiber-to-the-Home Council, & Telecommunications Industry Association. (2004). U.S. optical
fiber communities - 2004. Retrieved February 22, 2005, from
http://www. ftthcouncil.org/dbfiles/techexchange/OptFiberC ommunList1 0-20-04--
FINAL.pdf
Kelic, A. (2003). Networking technology adoption: system dynamics modeling of fiber-to-the-
home, technology, standards, and economies of scale. from
http://mphotonics.mit.edu/about_mphc/studentprojects.php
Keto, A. (2004). Everyone should have access to broadband, Bush says. Retrieved April 3,
2004, from http://online.wsj.com/article/0,SB 108032728080366544,00.html
North American Numbering Plan Association. (2004). Central office code assignment records.
Retrieved June 3, 2004, from http://www.nanpa.com/reports/reports_cocodes_assign.html
Warren Communications News Inc. (2004). The television & cable factbook. Retrieved August
11, 2004, from http://www.warren-news.com/factbook.htm
