***WAVE Comments
A foundation view for how home networking could emerge based on
my probing on the show floor.
As with last year, the move to digital CE products is what is
driving the industry at CES. Thus, the CEA seeks to promote all
forms of technology that will promote this growth. Certainly the
PC is a part of this but the migration of the established PC
players such as Intel, Microsoft, HP and Compaq to be more
"consumer ready" is very important factor in bringing excitement,
competition and new product concepts. Certainly the one area that
represents a different approach is home networking. The PC and IT
networking industries have begun this migration with the use of
802.11b in the home as workers bring notebooks home. However, the
home is not the corporate IT department and to understand the
form of home networking one must first visualize how the home
network may emerge.
Phase 1 - Confusion
In this phase there are many proposals for networking but none
that work, in terms of large market development. Included are
HomeRF, Wi-Fi, HomePNA, Powerline, and Structured Wiring. One
of the fundamental reasons is that these are derived from
legacy concepts of networking and these are exclusively driven
by what works in the enterprise environment not the home
environment. One of the fundamental issues is that no one
understands why consumers would use a home network. Thus, the
role of quality of data, bandwidth, ease-of-use, latency and
jitter are not fully understood. Further, there is a basic
market axiom often ignored - consumers do not pay for
infrastructure.
Phase 2 - Office Migration into the Home
This is the classic early adopter environment and includes
both structured wiring and 802.11b. In many respects we are in
this phase now with early indications of a movement to the
next phase. This phase is driven by the desire of the PC
plentiful to network multiple PCs in the home. It is very
important to note that the trends here are very unlikely to be
representative of the eventual market.
Phase 3 - Broadband Enters the Home
Now the market has transitioned driven by the value of
external connectivity. In the early part of this phase it is
only an extension of Phase 2 because the earlier adopters want
to have multiple PCs connected to the Internet. However,
significant barriers are erected such as banning multiple PCs
on at one time and no home servers. Another fundamental issue
is that broadband today is only broadband in that it is better
than dial up modems. No one with any technical sophistication
would call 1Mb broadband. In spite of these myopic structures
in the market it does represent a major step forward. Within
this phase is the adoption of residential gateways or smart
set top boxes. That is, electronic boxes which implement home
networking and external connectivity at the same time. There
are crude attempts with HomePNA but these will ultimately be
doomed. Over and over the market is turning to wireless as it
is the only assured means of connecting every home with an
infrastructure that is embedded in the equipment and thus is
controllable by what is sold and not what exists in any
unknown home environment.
Yet, again this phase is not driven by what the consumer wants
other than the early adopter. Lip service is given to
streaming content but the implications of this are not fully
understood, including technically, the market drivers
including costs and what consumers will support. Further,
because the US is behind in broadband penetration we believe
that many of the important developments in this phase will not
happen in the US.
At least in the US, this phase is further complicated by the
transport - DSL or cable. We have dealt at length with these
issues and this is not the forum for its rehashing. From a
standards perspective it is important to note that cable is
certainly a leader with OpenCable and its Cable Home
initiatives, which build on DOCSIS.
Bottom line from a home networking perspective wireless is
taking hold in this Phase 3. That is, in Europe Bluetooth is
playing a strong role while it is 802.11b in the home here in
the US. Yet, we are in the most early stages in this phase.
Phase 4 - Finding Something Useful to Network
It is only here that the mass market consumer plays a role.
Certainly the product today that stimulates this is the PVR,
however, it is not networked. This changed with the RePlayTV
4000 that has created a firestorm with the broadcasters and
Sonicblue locked in a court battle. There are product concepts
for media servers in the home, audio servers, photo storage
devices and many other home centric and network reliant
products. Network gaming, such as being promoted by Xbox, are
a part of this home networking drive but this is a small part
of the market given the narrow demographic.
In many respects Phase 4 is Phase 1 all over again. That is,
there is near total confusion in the product space. Many
proposals and uncertainly with what will stick. What is
different is that the product focus is now on what consumers
will find useful. In spite of all marketing hype this phase
has no meaning until products ship, consumers buy and trends
are established. To assume that market forecasts have any
relevance is folly. This is where Sony's "build it, sell it,
if it fails, dump it" is the best way to understand the
market. Few other companies really get this imperative.
This phase has one massive granite bolder in the way - fair
use, DRM and the like. Over and over in our conversations no
one wants to bring this up when speaking lofty words about the
potential of the products in this phase. But when the issue is
raised they rapidly fall back to - this is a BIG issue.
There is another very important element in this phase and that
is the embedding of home networking in products, at least in
the early phases. When networking products together it is
important that products will have this. Just as when 4 head
VCRs became the norm there was no turning back. However, this
early integration of networking into many products may or may
not include standards. Here I am not taking about Layer 1 or 2
standards but above. How do home devices communicate which is
meaningful in terms of what is being shared and how the
consumer will use it? In many respects, this is where the
proprietary development of the market will take place. Sony
did this on 1394 with its control protocols and not at the
transport layer.
Phase 5 - Integrated Networking
This is best described by analogy - the analog being the CAT 5
structure in the enterprise. Plug it in and it works. However,
what form this will take in the home is very different. That
is, there will be product threads, possibly started by only
two classes of devices, that are usefully connected. The
Microsoft Freestyle has a view of this so does Sony but these
are only market place guesses. It could well be that startup
companies may play a role here. It will only take a concept to
catch on which is reliant on networking that will drive a
major adoption of networking. To take this to the absurd the
value add may be due to functional integration such as timed
snack preparation in the kitchen during sporting events on
television. Thus, another factor is that these must be home
centric. The users, in the mass market, could care less what
the underlying OS is but on what is done for the consumer in
the home. We are a long way away from Phase 5. Not only is the
technology not here, but the mental agility to think in these
terms has not been reached.
Now we will shift to the standards side. Our biases here are
clear - only one technology counts for home networking -
wireless. Keep in mind that this is only at Layer 1 and 2 and
some of the higher layer issues, such as in Layer 5, are well off
in the horizon of the market. Here is our top-level assessment of
the market, as picked up at CES.
802.11g
Will supplant 802.11b - standard may not happen until Q3
2002 and if so, products not until Q1 2003; Atheros does
not want g approved soon - g cuts back on their market
which is based on 802.11a. There is a lot of "standards
politics" going on for the voting to approve this. 802.11g
changes the players and opens up the market well beyond b
today. Established players not in a position to quickly
leverage g will want to slow roll it.
802.11 h
Merged a and HyperLan2 (a is illegal in Europe). 802.11a
may launch in the US and Japan but it is not a worldwide
standard which will truncate is acceptability. Europe is
pushing HyperLAN2 and h is the means to integrate the two.
The potential exists with h that more than one supplier
will emerge. Today, Atheros is the only silicon company
with 802.11a.
802.11i
This is security. In spite of the claims of some it must be
implemented in hardware and this delays the market due to
the complexity.
802.11e
QoS and seen as critical. However, at present there are two
proposals on the table--one for data and the other for
isochroous content. If two proposals go forward this
considerably complicates the silicon and will likely hold
up time to market.
802.11d
Interoperable Access Point Protocol - applies to a and b,
Channel identification for Worldwide use. This is seen as
very important to the creating of a worldwide market. Under
this protocol a node queries the access point for the
channel structure it uses. With identification the node and
access point automatically configure the network to comply
with country standards.
Converged implementations are the direction of the future. We
have noted that Mobilian has implemented Bluetooth and 802.11b
in a chip set, there is no doubt that integration of multiple
wireless standards is the way the market will go. In a
parallel with Ethernet, where there was 10BaseT, then 10/100
BaseT and now 10/100/1000 BaseT; the legacy has not been left
off the network. With the building presence of 802.11b the
future standards cannot ignore the installed base. Thus, the
problem is placed on the silicon and with the complexity of
802.11, including that outlined above, forces multiple
standard support. We are very early in this process. Bluetooth
is likely to become a part of this converged silicon. Thus, in
terms of home networks, Bluetooth is a component for PAN
support and personal devices. Makes sense.
Today there is only one wireless standard that means anything
- b. Simple - this is all that is shipping in volume. However,
its use in the home is only driven by the early adopters. To
accomplish the transition to Phase 3 it must be more stable.
That is, the manufacturers of set top boxes will only use
technology that is stable and can go into millions of homes.
Even 802.,11b is not there yet. It could well be that only
with more standards maturation, as outlined above, will 802.11
emerge in Phase 4. Note I left off which version and in a
converged silicon environment the standards supported are
immaterial.
It was also agreed that a has the potential to play a much
larger role in the home than b. The reason is quite simple -
going to Phase 4 and many of the uses in this phase require
bandwidth. However, the end game is really 802.11h not a.
All of this points out the obvious:
The digital home is the center of very complex and
sophisticated technology; and
None of the underlying sophistication can surface to the
consumer.
This latter factor may well be just as difficult to accomplish as
the former.
Wave Issue 0204 2/08/02 Article 2-02