xDSL
Summit 2003
by John Latta
Wave Issue 0325 7/2803
June 16-18
Nice, France
Nothing makes it more evident that the US lives in our
own
telecommunications cocoon than a trip into market diversity. This
is certainly the case here in Europe at the xDSL Summit of 2003.
It was the same last year also. This is another excellent
conference put on by IIR.
www.iir-conferences.com
The US broadband market is very narrowly focused compared
to that
present here in Europe. Here regulation varies from encouragement
of the challengers to unabashed bias to the incumbents, such as
in Germany and France. Cable penetration ranges from high, where
it is an effective broadband challenger, to non-existent. Some
countries are well behind in broadband penetration, such as
Greece, to very progressive, such as Estonia. Certainly one of
the most interesting aspects of the broadband market dynamics are
the variations in how competition is handled. In many respects
this has reduced to the gap between the wholesale and retail
rates for broadband access in the local loop. If broadband
wholesale rates are low, competition is fostered and at the same
time puts pressure on the incumbents to compete. Yet, as we saw
in many of the cases today - low wholesale rates likely mean that
no profits are being made, especially in the lowest cost tier.
The impact is that the operators are seeking means to raise ARPU
and, of course, much of the discussion centers on broadband
applications. Unfortunately, the Europeans have nothing new to
add in the worldwide quest for applications.
The DSL Forum has as its goal 20% penetration of the world's
phone lines. They predict that this will be achieved by the end
of 2005. Point-topic, who actually compiles the DSL numbers for
the DSL Forum, in a separate presentation, predicted that there
would be 500m broadband lines by 2009. With approximately 1b
wired phone lines in the world these predictions all indicate a
fundamental shift in the telecommunications infrastructure. The
transition is as significant as the change over of the backbone
of telecommunications from switched to IP based. The issue today
is that the foundation for this shift is Internet connectivity
largely based on always on and faster. The demand side will have
to be greater if the impact of the shift is to be as important as
these announcements apply.
Tim Johnson - The Market Growth Opportunities for DSL
The growth in DSL lines was 90% in 2002. Japan is still
growing rapidly but Korea is reaching saturation. China is
expected to be a major player shortly. In 2002 Asia Pacific
lines at 16.4m exceeded the 11.3m in Western Europe and
North America had only 8.7m (DSL only). This was the year
that Western Europe passed the US installed lines.
On an individual country basis Japan is #1, with the US
second and South Korea third. Germany is #4 but
approximately 1/2 that of South Korea.
The situation changes dramatically when the countries are
ranked based on penetration. Again Korea is #1, with Taiwan
and Hong Kong following. The fourth country is Denmark. The
US is not in the top 10.
Point-topic is quite bullish on 1GE. They predict nearly
200m lines buy 2009.
In Europe, on a per capita basis, the ranking goes to
Sweden, Belgium, Denmark, Iceland, Netherlands, Finland,
Switzerland and Austria. This emphasis on the northern
latitude countries was jokingly referred to being related to
cabin fever - needing something to in dark winter nights.
It was interesting also that Iceland has no cable while in
The Netherlands the cable penetration is near 60%, as a
supplier of broadband.
A surprise in the grouping of the European countries is
Portugal and Estonia, being ranked 13 and 14 in terms of per
capita penetration. Portugal has 80% cable penetration while
Estonia has none - DSL only.
One of the contradictions was that a success factor for
broadband is regulation favoring the incumbent. This is
certainly the case in Germany and France.
New insights were provided when the broadband penetration
was computed based on GDP per person. In this measure
Estonia was ahead nearly a factor of 2 above Belgium.
Germany came in 12th.
As a follow up, it was shown that there is a high
correlation between GDP per person and broadband
penetration. This analysis was used as the starting point to
assess why are some countries much more advanced in the
deployment of broadband than others? This was most
insightful in showing the various reasons for the success or
failure of a country to accomplish broadband penetration.
Factors included:
- Government pressure for broadband adoption;
- Competition with cable;
- Unbundling or deregulation;
- Dominant incumbent carrier; and
- Established use of dial-up.
During the panel discussion at the end of the first day
the
issue came up - why is the US so far behind the rest of the
world in broadband penetration and related factors. In
response, Tim Johnson offered the following:
- For a long time the US was a leader in technology no
matter what is was. This is no longer the case.
Broadband is a good example. The Telecommunications Act
of 1996 was a failure. There are massive forces that
will block any initiative by taking any decision to the
courts to block action.
Sacha Lindecke - DSL Forum
The criteria the forum uses for Mass Market Penetration
is
20% of the phone lines in a country. Only Korea has achieved
this and the next country in line is Taiwan.
A strong push was made for DSL Home.
The following technical initiatives were outlined as
important to make DSL more competitive and to achieve mass
market status.
- Extend Reach of DSL
-- DSL Anywhere White Paper, ADSL2+
- Simply Modem Installation
-- TR37, TR 38
- Streamline and Automate Provider Order Flow
-- TR 38, TR52, TR53 and TR54
- Expand DSL Options to meet speed and traffic needs
-- ADSL, ADLS2+, SHDSL and VDSL
- Customer equipment Choice via Interoperability
-- TR-048, WT072 and WT079
It was stated that the focus today is now on home
networking. This includes initiatives in the following
areas:
- Security for Privacy and Protection
- LAN Side SDL CPE Configuration Specifications
- CPE I/F & System Configuration (TR-07 Update)
- Dual Port Router Requirements
- WAN Side DSL CPE Management Specs.
- TR-037 Autoconfiguration Update and
- Architecture to delivery enhanced services.
The broadband candidate applications:
- Online gaming
- Video conferencing
- Audio Streaming and downloads
- Video Streaming and downloads
- Real time video
- VPN and
- Remote security
Neil Quarmby - Current Developments DSL Technology (TI)
If anyone thought that 802.11 standards were complex one
only has to look at the ITU efforts on xDSL. Current
activities include:
- G.992.1 - ADSL (aka g.dmt)
- G.992.2 - Splitterless ADSL
- G.992.3 - ADSL2 (aka g,dmt.bis)
- G.992.4 - Splitterless ADSL2
- G.992.5 - ADSL2+ (aka g.adsl.plus)
- LDSL - Includes READSL2
- G.shsdl.bis - ESHDSL
- G.ploam - Physical Layer oam
- G.hs - handshake
- G.vdsl - VDSL
- G.selt - Single-Ended Line Test
- G.bond - Multi-Link DSL Operation
For example, ADSL2 has up to 7 annexes which describe how
this standard will work with POTS, ISDN and the Japanese
version of ISDN - TCM ISDN.
READSL2 - Reach Extended ADSL2 is to get at up to an
additional 1,000' from ADSL2. The rate is only 192kb/s
downstream and 96kb/s upstream. However, the best case
certainly appeared to be a special case and in other
conditions the reach was far less.
Some of the practical deployment challenges were outlined:
- Very high density cables at the CO - due to thin
conductors
- Many joints between different cable types which
result in impedance changes
- Mixed underground and overhead runs - opportunity for
RFI ingress and egress
- Some unterminated cable spurs due to bridge taps
Dynamic Adaptive Equalization was outlined as having a
major
advantage of expanding the service area reach.
Using a line from the cable book, in order to improve the
performance of DSL is was outlined how a hybrid approach
with fiber to the cabinet and then DSL under 2/5Km could
enable triple play, using ADSL2 or ADSL2+.
ADSL2 will begin deployment at the end of 2003.
2Wire - Landing in Europe
John Marshall, VP of Marketing, 2Wire, came from the offices
in
California to give the presentation. He began with:
"With our recent round of funding 2Wire is coming
to the
European market."
To make his point John showed recent compelling results
from a
" North American ILEC" which included:
- Consumers with only a DSL modem are 3X more likely to
cancel or switch broadband service than one with a 2Wire
gateway;
- Churn of those with 2Wire gateways is 71% lower than those
with DSL modems; and
- The take rate was 80% for applications when the CPE, i.e.,
integrated gateway, was included at a $10/month charge.
- "The price for our integrated gateway is more expensive
than a DSL modem but this device brings in more revenue for
the operator." As a result, SBC only charges the consumer
$50 for the 2Wire gateway and BellSouth provides it for
free. 2Wire is seeking to do in the Europe what it has
accomplished in the US with its DSL networking customers.
Some other points:
2Wire's business is enabling home networks with its
integrated residential gateway. Its recent round of funding
will support the entry of the company into Europe.
The company now has 142 permanent employees and 119
temporary, and the temporary are support personnel in call
centers.
The company claims the following:
- First DSL residential gateway - Fall 2000
- First Wi-Fi DSL residential gateway - Summer 2001
- First Broadband Applications and Backend Management
Platform - Summer 2002
- First to ship system on a chip gateway - Spring 2003
The 2Wire mission is:
- Develop residential gateways and backend management
systems as a service platform;
- Partner with the dominant DSL service providers to
drive home networking as a core service offering; and
- Expand opportunity through appliances and services.
The challenge home networking faces today is the last room.
That is getting to the any room in the home and always
provided broadband support. Further, we found it important
to provide support for many in-home communications
standards: HomePlug, PHNA, CAT 5 and wireless.
- 802.11g only provides another 40' to 50' and this is marginal
to solve the last room issue.
The ARPU opportunity is the home network and services
provided over broadband.
The Residential Gateway is at the center of a digital
lifestyle in the home that includes:
- PC;
- Voice and Video communications;
- Gaming;
- Media adapters;
- Home Security; and
- Home Control.
The 2Wire residential gateway is at the center of key
consumer offers by:
- SBC
- Bell South and
- BT
John then showed results which were not in the slide
handout. It is assumed that all these results apply with the
2Wire gateway and the survey results most likely came from
either BellSouth or SBC:
- The customer satisfaction of an intelligent gateway
is 26% higher than a DSL modem and 18% higher than a
DSL router.
- When compared to customers who have an intelligent
gateway those with a DSL modem are 3X more likely to
switch providers and 2X more likely of those that have
a DSL router;
- Measure churn was 71% for those that had an
intelligent gateway compared to those with a DSL modem
and 34% lower than those with a DSL router.
- Those that have an intelligent gateway:
-- 10% will buy bundles of CPE and data services
which are prices at $10/month;
-- 80% will buy applications where the CPE is
included at $10/month; and
-- 20% will purchase discrete services priced from
$2/month.
John stated that 2Wire will be announcing extensions to
the
gateway which will support home control and home monitoring.
There is a dilemma in retail. To show this John showed
a
picture of a retail shelf with many home networking
products. His comment was "Service providers need to take
control in the home to avoid the retail dilemma."
It was stated that some of the advantages of the intelligent
gateway is that:
- There is a superior installation experience;
- Future proof;
- Provides gateway management and reporting;
- Supports Broadband Applications; and
- It enables turnkey customer care/support.
The WAVE Report asked the question:
You state this is a future proof product while at the same
time imply many other potential functions such as a voice
gateway for SIP and home video monitoring. How is it
possible to design a product which will support so many
unknown future requirements and does this not imply a
processor which is much more powerful that you need today?
Answer:
Yes, we much over specify the product including the
microprocessor and equally as important the memory. We find
that we must go to 32MB.
Kingston Communications
Kingston Communications turned a broadband mind flip. This
is an
ILEC in Hull, England. They have a UK presence, including long
haul fiber, but the home market is the town of Hull on the
Eastern Coast. What Kingston has created is a delivery platform,
KIT, which uses DSL for transport. Theirs is a 5MB/s pipe which
is an operational and also experimental ASP delivery platform.
The primary use for KIT is video delivery. They are at the center
of the digital divide with low income users, many who cannot
afford PCs. Thus the delivery is television centric not PC
centric. When I asked Mike Alexander, Broadband Service
Development Manager, can KIT users get access to the 5Mb/s pipe
his response was simple - they would not know what to do with it.
What is striking about KIT is that:
- The platform supports content created by users, such
as
personal photos and videos, including storage and publishing
on the KIT platform;
- All video content can be on demand;
- Digital television is the primary service; and
- Kingston Communications encourages ASPs and the
development of new services.
The lists of firsts include:
- First commercially deployed full service broadband TV
using ADSL
- Largest commercially deployed VoD system
-- 35 nCube video services capable of 7,000
simultaneous streams
-- Video capacity of 7,000 hours
- First broadcast VoD interactive advertisement
- UKs first interactive drama and
- BBCs first TV based VoD news service.
The KIT platform is based on:
- IP
- HTML and JavaScript
- JPEG, GIF and PNG and
- MPEG and RTSP.
- The result is that they believe content development and
delivery is more cost effective than any other digital TV
platform.
There are 70 channels. Real time video is encoded on the
fly at
4.3Mb/s and off line is at 2Mb/s. There is full VCR functionality
for all video content.
A video was shown of the service and it was impressive
in the
depth of what is offered.
Users can download pictures and personal histories. They
can also
create videos and download these. These personal content products
are 2nd only to VoD news in terms of usage in KIT.
Hull is in the center of the digital divide. The PC penetration
rate is low but the details were not stated. However, KIT offers
PC availability as a service using Citrix thin client. This
allows any Windows or Unix applications to be used on the
television.
Kingston Communications specifically seeks out ASPs. In
order to
do this they work on a revenue sharing arrangement in order to
share the risk with those doing new application development.
The WAVE Report asked Mike afterwards what exactly are
the
subscriber numbers for KIT. It had 10,000 but they recently did a
" churn" and deleted some 3,000 that were not paying the bills -
assuming they could not afford the service in the poor region of
Hull.
Swisscom Fixnet Wholesale - Broadband Realities
Hanneke Gerritsen gave one of the most challenging presentations
on the role of wholesale services, competition and innovation.
Swisscom has been under fire. Cable was first to market, the
company has an infrastructure poorly suited to broadband and by
regulation it can only provide wholesale broadband access. All
customer contact is via independent ISPs with the Swisscom Fixnet
wholesale model. In the first year of service, 2001, there were
only 40,000 subscribers and by Q1 2003 this has risen to 315,000
subscribers. The ISPs are where the customer innovation takes
place and there are approximately 24 of them. The thrust of the
efforts at Swisscom is to create a new product which will drive
demand. Initial efforts at content distribution by their ISPs
have been a disappointment. Swisscom is developing a service
model which is very ASP like with the expectation of developing
content or application partners. This has significant potential
to encourage innovation in service delivery. Yet, when asked this
is difficult due to the out dated billing systems and poor
physical plant. The reality is that cable is out innovating them.
This is a PTT that has no retail broadband presence. It
is
at the opposite end of the position that the ILECs in the US
are seeking. Thus, Swisscom is dependent on its ISPs to make
the business happen and it is solely a wholesale operator.
There are two offerings: Private (consumers) and Business.
The private broadband rates vary from 30 to euros and the
business rates from 130 to 300 euros. The variation in rates
is based on bandwidth. Thus, the incentive at Swisscom is to
sell more bandwidth or value added services, content in
their terms. The major business offering is VPN while the
plans for Private are new content products. As they look the
future there are two options:
Option 1 - Status Quo
Continue to drive the penetration. However, there
are pressures on the bandwidth price and thus to
lower the ARPU. The downside, with out additional
revenue streams is lower revenue and
profitability.
Option 2 - Innovation
The emphasis is on new service oriented business
models. The expectation is that this will permit
higher ARPU.
The view is that these products will demand more bandwidth
and QoS. The down side is that the complex protocol
architecture in the Swisscom network limits what can be
done. Thus, they are faced with CapEx expenditures. Given
that Option 2 was about Innovation I asked are the cable
competitors out-innovating your company. The answer was a
simple yes.
The diagram for the new service model based on Option 2
was
the creation of a services business which they called
" Content over Partner." This is very similar to an ASP model
which has a third company seeking value added services.
Swisscom agreed when asked. There are closer parallels with
DoCoMo's model for ASPs. However, the Swisscom emphasis on
the ISPs and the only means to get to the end customer seems
limiting. They partially agreed but emphasized the role of
the relationship with the ISPs.
Iceland Telecom as a Microcosm of Telecommunications
The WAVE Report met with two attendees from Iceland Telecom,
one
responsible for its broadband strategic planning. In spite of the
country having only a population of 290,000 we were impressed by
the state of the technology and market dynamics. Iceland
telecommunications has been deregulated and has a limited form of
structural separation. For example, the telecommunications access
provider operates independently of the ISP. The Iceland Telecom
ISP pays the same wholesale rate for broadband access as does any
other ISP. This again puts pressure on Iceland Telecom to make
money on each service it provides. However, the lowest tier of
service for $33 loses money. Iceland Telecom, came to this
conference, in part, to seek out new applications for broadband.
DSL is now available to 90% of the homes passed by Iceland
Telecom phone services. It is highly unlikely that this will
be extended - to do so would be unprofitable. However,
virtually 100% of the homes have access to ISDN but the
bandwidth is limited to 128Kb. Because of the rural areas
where Iceland Telecom serves some homes cost up to $10,000
to even provide this service.
The ISP and the carrier portion which provides DSL must
operate independently and although the financial data is not
independently reported, the government regulators have full
access. Competition at the local loop is encouraged and
Iceland Telecom pays the same rate for a local loop as a
competitor does.
Since 1995 in Reykjavik a plan for HFC has been embarked
upon. Currently 50% of the city is served by this
infrastructure.
In contradistinction with the US, there is no T1 - T3
service gap. Iceland Telecom has many offerings in this rate
range.
Iceland Telecom has two peering arrangements - US with
UUNet
and Europe with Cable and Wireless. Both are at 155Mb/s.
Although the DSL fee structure is "all you can eat" the
exception is for both inbound and outbound traffic via the
peering points, which is charged above a certain limit.
Currently they are running at about 15% of capacity of these
links.
Iceland Telecom has an ATM backbone network but this is
being replaced with an all IP MPLS network. As we learned
with AT&T and the similar plans they discussed at Supercomm,
this change over will be transparent.
When the ADSL service was developed it was decided as policy
not to block any ports. Thus, users would not have
restrictions to the Internet.
The rates for ADSL vary from $33 to $800 per month and
this
goes from 256Kb/s to 1.5Mb/s. At the lowest rate they cannot
make any money. The major ISP competitor offers the mid-tier
service at 512Kb/s at the cost of the lowest tier at Iceland
Telecom - $33. They made a decision not to match this given
the losses already being incurred at the lowest tier.
Consumers buy their own ADSL modem and self install.
Users can purchase a fixed IP address and even do web site
hosting. Limitations of this are the costs of in-bound
traffic from the peering points. This can cause a
significant rise in the monthly bill.
Iceland Telecom has been experimenting with various bundles.
One at the mid to upper tier rates includes a bundled web
cam and this has been quite successful with a 30% take rate.
Another bundle under consideration is similar to what KT has
been offering. That is, home networking based on wireless
with a tie to hot spot use in the same pricing bundle.
Some companies have tried VoIP using broadband to the home.
This failed because the phone rates are so low as to make
any marginal value of VoIP as uneconomic.
There remains speculation that content over broadband will
drive usage. There are two issues, if one assumes that the
driving content is video. First, video rental stores are
everywhere, most within a short walk of apartments and
homes, that providing VoD could well be marginal against the
walk in competition. Secondly, getting competitive content,
against what is available from the video stores, is very
difficult. Iceland is sufficiently small that the cost of
the attorney fees to cut a deal, if one could be cut, would
swamp near term revenue streams.
Iceland Telecom has servers at the ISP for low latency
game
play, which they feel is a competitive advantage. One of the
fun aspects to this are the Doom Beer parties where player
bring over notebooks, set up a LAN, connect it to the
network with DSL and game play is fast using the Iceland
Telecom servers.
One of the major problems in telecommunications in the
industry and at Iceland Telecom is that the technology
changes so fast that it is not possible to write off the
investment before the technology is obsolete.
Aliant - Mile a Minute Presentation
Paul Hanson might be called a presentation star. He even
had a
sponsor for the presentation - Nortel Networks. The topic was on
the SME market but this was only a part of this talk. Aliant, the
company he works for, is the incumbent carrier in the Atlantic
region of Canada. It was combined from four other carriers in
1999, has sales of $2.6b can, and 10,000 employees. The company
has:
- Local and Long Distance
- Wireless
- Internet
- E-commerce
- Innovatia - eLearning
- Business telecommunications services - worldwide
The company has:
- 1.5m wireline customers
- 477,000 wireless customers
- 300,000 Internet customers
Paul's strength as a speaker is his ability to cruise through
information dense slides, tell good stories and flood the
listener. It was a good ride.
- A killer in the consumer market is churn. Cox has found
the following:
-- Video only - 1.5%
-- Video + data - 1.0%
-- Data + Voice - 1.1%
-- Video + Data + Voice - .7%
- The greatest service profitability is nothing to do with
broadband. A very interesting chart was given that showed a
90m film creates a revenue of $.01 per MB while a SMS
Ringtone is responsible for $13,981/MB.
- There is an emerging world wide communications market
evolving which entails connecting remote offices, assumed
small, with ADSL, since this has become very widespread. At
each office is a router with VPN. Thus, the whole network is
established as a VPN. The customers only like VPN which have
a hardware box on the individual premise.
- Paul showed "war chalking" icons for wireless
access.
These include: Open Node (no SSID), Closed node and WEP
Node.
The WAVE Report asked Paul: Have you worked in Broadband
Applications? Paul replied:
Yes. We have not made much progress. In spite of extensive
consumer research there are no good indications of what will
drive the market. The published research is not a good
indicator. Our experience has shown that the public is a
poor predictor of its intent. We have even gone so far as no
longer run free trials. What the public does in a trial is
not a good indicator unless they pay for it.
It is our experience that the business market, even for
ADSL, is much more predictable and provides a greater return
on the investment.
EFMA - Ethernet First Mile Alliance
Craig Easley came to provide an overview of the alliance
and
sought to create a chapter here in Europe.
There is the usual split between standards and marketing.
The standard for EFM is under 802.3ah while the marketing
issues are under the EFMA.
The goals of the IEEE standards effort is to:
- To develop an Ethernet standard to support these
access topologies:
-- Point to multipoint (P2MP) on optical fiber
(EPON)
-- Point to Point (P2P)
-- Point to Point (P2P) over copper (or EFMCu)
--- Short Reach/Very High Speed
--- Long Reach/ High Speed
-- Operations Administration and Maintenance (OAM)
The controversy has centered on the EFMCu. There continues
to be a debate if Cu plant makes any sense for Ethernet. The
short reach/very high speed option is restricted to only
750'.The operators, North American ILECs, wanted a better
solution and is resulted in the Long Reach/High Speed
option. The long range is only 2Mb/s and with a range of
2,700m. There is an option for the bonding of multiple
pairs.
The P2P standard supports both 100Mb/s and 1G/s transport
on
both single and dual fibers. The reach is 10km and will
operate outdoors - the equipment is specified to operate in
a range from -40 C to +85C. The PON version supports up to a
16:1 split ratio. An interesting aspect of EFM is the
ability to combine the access topologies into hybrids. That
is Fiber P2P can feed EFMCu or Fiber P2P. Under these
conditions the OAM should pass thru the hybrid topology
tree.
It was interesting that the IEEE would not decide the "Line
Code Issue." That is, what modulation is to be used in the
transport. The options are DMT (Discrete Mult-Tone) or QAM.
This was handed over to the ITU and a decision is to be made
shortly.
The standard will close on July and go for a working group
ballot at the same time. It is expected that the final
standard will be released in June 2004.
The WAVE Report asked - did the BPON intent statement by
the
ILECs include EFM. No, as this is not yet a standard. The
ILECs went with another standard and this is the only one
they could use. However, it is expected that the process
outlined by the ILECs will take a long time and the role
that EFM could play remains TBD.
***WAVE Comments
The struggle for applications and the implied link with
bandwidth
is a continuing quandary. There are two ways to consider the
basis for driving the value of broadband delivery up--from an
application perspective and from a bandwidth view point.
Applications
The fundamental issue here is:
- What services will consumers find that leverage a
digital infrastructure?
The more we listen to the arguments about applications
we
wonder if the business model will ultimately center on ASPs
- Application Service Providers. That is, companies that do
nothing but seek out specific vertical applications.
Certainly DoCoMo's i-mode experience is representative.
The experimentation at Swisscom is very forward looking.
But
this is mired in the weakness of their OSS and the network
itself for the delivery of such applications. Most
importantly the assumption is made that the ASP choice is
not a la carte but is gated by the ISP. Thus, to say there
is an ASP model is quite different than delivering on it.
It is highly unlikely that ASPs useful in a home will be
stovepipe. A single company will not be able to provide a
stand alone application. Thus, partnerships will be
required. The reason for this is that most likely high value
home applications will require multiple parties--for example
home and school interaction.
We are faced then with multiple complex issues:
- Applications are not clear;
- How to deliver them and the role of the broadband
infrastructure is also lacking; and
- The business models and business development means is
as cloudy.
Broadband Use
This perspective is one of the technology first and
applications are expected to follow. This tactic is what
most telecommunications companies take. It is natural for
them because bandwidth is the metric which defines a key
attribute of the network.
If one asks - what would drive bandwidth usage up we see
three factors:
1) Large number of devices whose aggregate forces
bandwidth demand;
Many small, low-bandwidth devices would aggregate
to a total demand for large bandwidth. This is
consistent with Metcalf's observation that the
value of the network goes up as the square of the
connected devices. We assume the domain is in the
home. Further, this also implies that the nodes
would disappear - the consumer would just take the
network and its functions for granted.
2) Disaggregated computing elements, which require
broadband to interact;
This implies that the home is an element in a
computing or IT system. Such a concept may or may
not be linked to the aggregate device proposition
above. However, for the consumer to rely on
storage external to the home for the conduct of
the affairs of the home implies considerable trust
of those in the home. Such trust relationships
exists with the family doctor, faith, and bank but
certainly not with the computer. This is
especially the case when the components are not
under the direct control of the consumer.
3) Content which drives bandwidth and this includes:
Video, especially that originated within the home;
and
Large file transfers.
The difficulty of a bandwidth centric view of the demand
side is that none of this is near term. Possibilities
include video in the home, baby monitoring, home security
and family communications.
The points to consider then are:
- In the near term equating bandwidth with high value added
services may be an over expectation.
- The real issues lies in finding out what consumers will
use and pay for.
-
I
|
