The Future of Networks
By Carlo Longino, Thu Oct 17 00:00:00 GMT 2002

Would-be successors are already lining up behind 3G


It seems inevitable in the technology
world that even before the next technology has arrived, the hype machine
and commentators have moved past it. The mobile industry is no exception
– what’s available today is never good enough, what tomorrow holds is
always better.

Perhaps it’s a way for companies to deal with
perpetual disappointment: “Oh, just wait for the next version.” Perhaps
it’s just a product of wishful thinking: “Wouldn’t it be cool if…”
Perhaps it’s just the course of technology.

But whatever the
reason, some companies are already looking past 3G, some around it and
others over it. So the question isn’t if mobile data will take off, but
with what technology. UMTS and WCDMA haven’t really emerged yet, and are
already being written off in some quarters. WLAN has exploded, but still
has yet to prove its viability for widespread networks.

There
are scads of new products and technologies that will emerge and
supersede, compete – or dovetail -- with WCDMA 3G networks. Some may
come into use in the short-term future, while others might come to be
known as 4 or 5G.

802.competition

The
most obvious alternative to 3G is WLAN. While pundit after pundit has
said the 802.11 standard’s popularity will negate the need for 3G, it
may pose as much of an opportunity for carriers as it does a
threat.

Public WLAN networks have stormed onto the scene,
offering a lot of benefits for consumers and providers alike. They’ve
whetted the public appetite for high-speed mobile data, and most of all,
they operate in free, unlicensed spectrum.

“Prophets follow free
resources,” says telecom analyst Seamus McAteer of the Zelos Group.
“WLAN is forcing everyone to re-examine their strategy, and is a
disruptive force that precedes wideband CDMA.”

Although WLAN
networks’ attractiveness and potential threat is clear – why would
consumers pay through the nose for relatively slower access when they
can cheaply and easily use much higher speeds --problems inherent in
the standard’s short-range signal and its lack of support for roaming
make its current incantation an unlikely threat to 3G.

“WLAN is
inherently unsuited for metropolitan-area networks,” McAteer says.
“It’s hard to see a viable near-term network nationally.” But products
are emerging that could remedy some of these problems.

Wireless
network equipment maker Proxim (which sells a third of the world’s
Wi-Fi equipment) has released souped-up 802.11b equipment that can send
its 11Mbps signal as far as 12 miles. The gear is relatively cheap, as
well, and is aimed at small ISPs looking to go wireless. A system sells
for between USD 2,000 and 6,000 depending on the level of equipment and
other options, and can serve up to 250 users.

While this
eliminates one problem – the short transmission distance – of WLAN, it
doesn’t solve two others. WLAN is intended for nomadic, rather than
mobile use; it supports a move-stop-communicate model, and can’t really
keep up if a user is traveling at much more than 5 miles an hour. More
importantly, it still doesn’t support roaming. But another company is
tackling that issue.

New Zealand startup RoamAD has developed
the world’s first cellular Wi-Fi system. It’s already built a
3-square-km network in the central business district of Auckland (and
plans to expand it to 100 square km) that provides users with ubiquitous
330 kbps access, nearing 3G’s initial theoretical throughput of 384kbps.
All this, the company claims, at 5% of the cost of a similar 3G network.
And it’s compatible with existing devices. Add a VoIP client, and it’s a
phone network, too.

RoamAD’s approach entails a coordinated
installation of network elements, and adjusting their access points’
identifiers so all users’ devices see is a cloud of access, rather than
individual APs. The company has also developed its own security system
based on IPSec rather than the creaky WEP standard, and its star-grid
configuration means that users don’t have to maintain line-of-sight
connections and there’s enough redundancy (users are in touch with at
least 4 access points at all times) to handle a large volume of
traffic.

Systems like these will enable nearly anyone to enter
the wireless ISP business and pose a minor threat to carriers. But the
ubiquitous nature of mobile-phone network coverage offers carriers a
major advantage, and makes these technologies a great supplement to 3G.


Meshing It Out

Another
hot area of development has been in mesh and ad-hoc networks. They’re
often referred to as “infrastuctureless” because they allow devices to
communicate directly with one another without going through a central
network. This architecture has lots of implications, because every
device on the network acts as a transmitter, not just a receiver. This
allows for quick and cheap network expansion, as well as organic growth
– not just in coverage area, but importantly, also in quality by adding
additional redundancy and connections to the network.

One
leader in this area has been US-based MeshNetworks . Their network
design relies on one central switching center and an access point, then
uses wireless routers the size of a large deck of playing cards to
communicate to devices on the network. Devices can communicate
peer-to-peer, create ad-hoc peer-to-peer groups, and make multiple
hops, if necessary, across other users and the wireless routers to
reach the network.

This also serves to increase throughput –
radio physics dictate that when power is constant, the only way to
increase data rate between transceivers is to decrease the distance
between them. The network can burst up to 6Mbps, and deliver sustained
rates between 1.5 and 2Mbps.

So if a user wants to communicate
with another user, they can reach them directly if in range. If not,
they can hop across other users and routers to reach them. This allows
for the quick and easy build-out of networks across city-wide areas.
The “seeding” necessary to build a city-wide or regional network is
cheap as well, making the economics extremely compelling.

“Our
cell sites cost a couple thousand dollars, not a couple hundred
thousand dollars,” says Rick Rontondo, the company’s vice president of
technical marketing. “And as you get more and more clients, they extend
the network as well. The third thing that is very different, is if you
go to an area that has no coverage, you still can form peer-to-peer
networking between groups of people.”

This last factor also
makes the networks attractive for public-safety use or in emergencies.
If a number of firefighters and paramedics descend on an area, they
could immediately communicate with one another without any
infrastructure. The company has already crafted such a network for one
US police department. Each police car has one of their transceivers,
allowing officers not only to access databases at high speeds, but also
allowing the central control – or any other officer – to observe other
officers’ situations thanks to a link to the cars’ dashboard-mounted
video cameras.

MeshNetworks also sees their products as
well-suited for telematics. They say the economics are compelling enough
for manufacturers to outfit cars with their devices, then connecting
those to a few access points around a city – perhaps located on the
roofs of that manufacturers’ dealers – would quickly create a high-speed
network. Their product isn’t motion-limited like WLAN; it will function
at up to 250 miles per hour. And roaming is no problem, either, as their
software can route packets in 5 milliseconds.

MeshNetworks’
system is radio-agnostic, meaning it can be engineered to work with most
any standard. But by its nature, it does away with cellular network
architecture, but again, add a VoIP client and a PSTN connection, and
it’s a voice network. That’s a secondary feature, however, as the system
was designed from the start for data.

“If I was going to build a
voice network, I’d build it like a cellular network if I was doing it
wireless,” Rotondo says. “But data? Haven’t we learned anything? We
tried doing it on the telephone network, and it failed miserably. What
stuns me is that you have these guys who are trying to build broadband
wireless networks using the same architecture Alexander Graham Bell
deployed 150 years ago..”

Flights of
Fancy


And there are even more ideas that
could come into play in the future, and paralleling the early days of
3G, they range from the impressive and realistic to the more,
uh…fanciful.

Flarion Technologies has taken a radio transmission
breakthrough, flash-OFDM (or orthogonal frequency division
multiplexing), first developed at Lucent’s Bell Labs. It, too, is a
data-based, all-IP system that makes no distinction between packets, be
they voice data or Web pages. It claims a latency of 35 milliseconds and
a top speed over 1Mbps. It also only uses 1.25 mHz of spectrum, a
quarter of the typical 3G network, and the company says it costs a tenth
of 3G to roll out per subscriber. Flarion’s technology is currently in
trials with US carrier Nextel and South Korea’s SK Telecom.

One
US company has an audacious plan to blanket the country with coverage
using weather balloons. Space Data plans to piggyback on the 50,000
National Weather Service balloons launched all over the country each
year with 6-pound repeaters. Seventy balloons are launched each day by
the NWS and fly at an altitude of 100,000 feet for about a day and a
half at a time.

The company plans to “launch” a nationwide
two-way text-messaging service as soon as next year, and says the
economics give it a good chance of success. The company’s CEO estimates
a wireless carrier would need at least 5,000 cell towers to cover the
10% of the continental US where 80% of the population lives. If the
carrier is spending an average of USD 1,000 per month to lease each
tower, that adds up to USD 60 million per year. Space Data will launch
50,000 of its repeaters at USD 300 a piece, so for only USD 15 million,
the company can cover the other 20% of the population and 90% of the
land mass.

That’s assuming no balloons are recovered – the
weather service manages to reclaim about 18 percent of its balloons,
but Space Data hopes to improve on that by adding GPS locators to its
repeaters.

Back in the labs, NTT DoCoMo has announced it’s
successfully tested a “4G” network that transmits at 100Mbps downstream
and 20Mbps up. It, however, requires a massive 100 mHz of bandwidth.
DoCoMo says the system uses variable signal frequency and orthogonal
frequency code division multiplexing technologies to cut down on severe
multipath interference generally associated with networks of that
bandwidth.

The company believes the necessary technologies for
commercial deployment could be developed by 2005, and launched by 2010.
That may be overly optimistic – or economically unfeasible – but it’s
further proof the future’s coming faster than you’d
think.

It's Future Week on
TheFeature! Check back daily for reports, analysis and in-depth
articles.




Carlo
Longino is a freelance writer based in Austin, Texas. His
previous experience includes work for The Wall Street Journal, Dow
Jones Newswires, and Hoover's Online.