CWI Cover Story. Backs to the future
totaltele.com
By turning their backs on market needs for low-cost mobile
data, the world's wireless operators could find themselves
rooted firmly in the past, along with the incumbent PTOs, as the
future passes them by. Bob Emmerson reports
Want to shake the telecomms establishment to its very
foundations? One way to do it is to think the unthinkable. How
do you do that? Well, it's remarkably easy. First of all, imagine
Internet access over a high-speed wireless loop. Now, add
the concept of packet-switched voice over mobile handsets.
Then, think of ISPs hooking into a global IP backbone to
become ITSP's and there you have it. Mobiles doing
carrier-class voice over IP (VoIP), and literally, pushing the
PTOs "out of the loop". A radical departure that leaves
incumbent telcos, and even some new carriers, facing the
wrong way as the rest of the world rushes on past them: and the point is, it could
happen.
Why? Because the world is crying out for higher-speed access to the the
Internet, and lots of telco-sponsored technologies simply cannot provide it.
For example, ISDN won't do. Admittedly, ISDN log-ons are quick, but its data
transfer rate isn't up to modern requirements and circuit-switching is the
technology of yesterday. PTOs based the ISDN B-channel rate of 64kbit/s on
the capabilities of transistors (yes, ISDN is that old) and then, in the sluggish way
that characterised the days of monopoly provision, took 20 years to bring it to
market.
Elsewhere, xDSL is a good technology, and its data rates are up to the job, but
apart from a few trials, and even fewer installations, it is not widely deployed.
Indeed, the entire technology might be another wasted opportunity for the global
telecomms community.
And remember fibre? Once upon a time it was going to be fibre to the home,
then it retreated a few metres to become fibre to the kerb. Then it retreated even
further back over the horizon, out of sight and almost out of mind. However,
fibre isn't entirely dead - the new carriers are fibre-cabling business communities
and cherry picking the PTO's best customers.
Coaxial cable is another fat pipe that could supply the necessary bandwidth. It
already goes into millions of premises and residences worldwide - a key benefit -
but the coaxial industry is fragmented and CATV technology had to be modified
to provide upstream data transmission capability.
While this wasn't like developing the Special Theory of Relativity, it took a
similar length of time. Cable modems, which use a parallel technology, have also
not yet been taken up in significant numbers.
And then there's Digital European Cordless Technology (DECT) which proved
to be yet another technology, and one with a potentially high channel rate that has
also failed to be implemented widely.
Controlling access to the last mile
What the world's users really need is a single, high-speed communications link
that does it all. Without that, there can never be universal access to true
multimedia. Incumbent PTOs have clearly let customers down and missed out on
an enormous market opportunity. They're losing business, some would say
deservedly, to the new carriers and IP backbone networks are bypassing the
international tolls, which is where most of their profits come from. But they still
own the local loop - the infamous 'last mile' that has turned out to have such a
stranglehold on the aspirations of new carriers and subscribers alike.
One possible solution could be multi-slot general packet radio service (GPRS).
This technology can be pushed all the way to 384kbit/s, is packet-switched
(which makes interfacing with IP incredibly easy) and can be used for both voice
and data. Furthermore, it seems likely that GPRS will soon make it into the
WAN arena.
This enormous opportunity is staring the GSM community in its corporate face,
so what are they doing about it? Not enough to satisfy the market, that's evident,
but action on the data side of the industry is picking up.
A single voice
IP is the lingua franca of data communications and voice is sure to become part
of that - it is only a matter of time. Today some 40% of the workforce of the
developed world are mobile, or soon will be, and many of them interact direct
with customers. Thus, the market for the wireless intranet exists, sales of the new
wireless information devices are going to take off and Wireless Access Protocol
(WAP) will create a global market for value-added services. So even with the
local loop excluded, now is the time to roll out GPRS. The trouble is, the
industry is trying to buck the market by touting high-speed circuit switched data
(HSCSD), using multimedia as the spin, when what is really needed is low-cost,
ubiquitous packet-switched mobile data. And in this day and age, the market will
win.
The virtues of N x 14.4kbit/s (GSM can move up to this figure from 9.6kbit/s)
are being extolled, although a download speed of 43.2kbit/s is the practical limit.
The reason for all this myopia is equally clear: HSCSD is a relatively easy
upgrade for the operators since it's done with software. GPRS is more complex
and new packet-switching hardware is needed.
The GSM industry has turned its back on mobile data for both the wireless
internet/ intranet as well as the local loop. Rather than accept the reality of the
emergence of a massive new market, it has concentrated on signing up more and
more voice subscribers. There is nothing wrong with such a strategy - except its
short-termism.
Danger up ahead
What is worrying is that the GSM operators, albeit for different reasons, seem to
be walking into the same trap that cost the PTOs so dear. GSM itself didn't
require a massive marketing push to make it a success. In fact, the technology's
popularity took the industry by surprise (with the notable exception of Nokia)
and GSM was hardly marketed at all.
Independent ICT consultant David Greetham told CI: "Back in 1991, when the
first GSM networks went operational, there was no coherent vision, nor was
there any master plan mapping the way to today's environment. No one
expected this level of success, it just happened. The handsets were huge, heavy
and incredibly expensive. Expectations were that a few thousand would be
carried around by senior executives in their limousines. That's why plans called
for only a few base stations to be deployed along main roads outside the cities. It
was thought that would be sufficient to carry all traffic. Separate networks had
been set up for mobile data, and the convergence of the two into small,
lightweight, easily portable, personal multimedia devices was regarded as being
proper to UMTS and the far distant future."
That same impaired vision is still evident today. The PTOs lost business, and
continue to do so because they ignored market needs and clung to their
monopoly status and privileges for as long as they could. Now many are paying
the price of such hubris.
What is so disconcerting is that the GSM operators - creatures from a different
age, with a different agenda, and surely, who should know better, seem
determined to follow the same route as the PTOs. This could be the road to ruin,
for, in trying to buck the market and determining that they know best, the GSM
operators are demonstrating that they are little more than short-term
opportunists, lacking the vision to plan for the long-term future.
Of course, GSM has come a long way from its original incarnation and now
routinely delivers many of the services that were envisioned for UMTS. Indeed,
the current target is for a so-called 'world phone', able to handle data rates of up
to 2Mbit/s.
However, the trouble is that the timetable for the introduction of 3G services is
subject to constant revision. Furthermore, given the stand-off between Europe
and the US over intellectual property rights and the global convergence of 3G
standards, the debate, (if we can describe a mud-slinging exercise as such),
seems likely to drag on for a year.
The big tease
As for the market, well it's not going to get worked up about new services until
they come on stream, and for operators to dangle before it the lure of 2Mbit/s
services without ever providing them, does nothing more than add insult to injury.
The injury being that the vast majority of networks provide just a 9.6kbit/s
circuit-switched connection. Only a handful of operators can provide 14.4kbit/s
in a handful of networks, and the much-vaunted multiple channel 14.4kbit/s
networks are only in their earliest stages.
The insult, meanwhile, is in the multimedia hype that surrounds HSCSD and the
implicit message is that the industry, rather than the subscriber, knows best.
Certainly, the new service will give operators something to promote, but the
reality is that, deep down, they are at a loss what to do. The dilemma is whether
to go for GPRS/EDGE, or wait for the arrival of 3G? A data rate of some
400kbit/s is clearly attractive, and anything higher may well be superfluous, given
the incidence of Internet bottlenecks.
The nub of the issue is whether 3G is a market demand or an industry construct -
and there is a strong feeling in some quarters that 3G is an attractive solution to
operators in the 900MHz band that are finding themselves embarrassingly short
of spectrum.
A clear path
Nonetheless, from both a technology and operator perspective a clear migration
path is discernible: 9.6kbit/s > 14.4kbit/s > HSCSD > GPRS > EDGE >
3G
The first two upgrades are relatively easy, and indeed moves are already afoot
provided there is enough spectrum. But after that, things are less clear. The
crucial question is that, if operators opt for GPRS, will they raise sufficient extra
revenue to justify investing in 3G? If they wait for 3G, they may be denied the
spectrum they need to implement it. (W-CDMA is planned for 3G, thus making
more shareable spectrum available. However, the migration to 3G also
necessitates a change in wireless interface and this would cost a great deal.)
The end-user perspective is also unclear. New devices would be needed for
each and every major upgrade and the market simply would not stand a
prolonged series of additional expenses for incremental benefit.
Until a few months ago, Nokia was vocal in backing the idea of a PC Phone
Card that could handle both HSCSD and GPRS, but has since been silent on the
matter. This reticence will almost certainly be enough to guarantee the failure of
N x 14.4kbit/s. The perceived difference simply isn't worth the subscriber paying
out for new devices. No, the market will wait for GPRS and pray for EDGE.
Meanwhile, Microsoft has been pushing GSM handset manufacturers to adopt
Windows CE, but Ericsson and Nokia have opted for Psion's 32-bit real-time
operating system instead, and now Motorola too has joined this camp.
Elsewhere, Qualcomm has been appealing against what it claims is a series of
IPR fouls, and that company's actions does cause one to wonder what might
have happened if this particular technological boot was filled by a US foot.
All the same, the two companies have come up with an appealing vision of a
wireless data future and WirelessKnowledge has been expanded into an alliance
that includes Compaq, Hewlett-Packard and AT&T Wireless. Microsoft's
agenda is clear: get Windows running on wireless thin clients and sell more NT
servers. The stated objective is to "develop a platform for a wide range of data
services".
Qualcomm's position is less obvious. It clearly has the requisite wireless
technology, but this move also seems to indicate that it is discarding its own
consumer software ambitions. It also has many products that compete with
Microsoft's Exchange mail server and Outlook e-mail and contact management
program. Qualcomm also clinched an earlier deal with 3Com. The PalmPilot
(arch rival to the MS CE camp) is combined with its digital wireless phone in a
product known as pdQ. This deal was obviously put together in a hurry.
WLL: an impossible dream?
The enlarged entry of the US computer community into the wireless arena
convinced onlookers that the idea of using wireless technology to enable
high-speed access wasn't as off the wall as previously thought. Essentially,
Microsoft has created a broad alliance, a vision, and a business model which has
generated a variety of opportunities, ranging from IP telephony and Web hosting
through to enhanced television and e-commerce.
The vision calls for a common infrastructure - for billing, security and
management - and a technology-neutral platform that accommodates the current
mess of incompatible wireless standards.
But no matter how broad the alliance, given the chaotic state of cellular services
in the us, it will take several years before anything worthwhile can be achieved
there. The situation in Europe and the rest of the GSM-centric world is very
different, but even here the industry continues to view the future through
technology glasses which give it a very rosy tint.
The truth is that no one yet has the ideal technology to enable relatively
high-speed Internet access and packet-switched voice on the same network -
but Europe is closer to achieving it than the US, or at least it would be if it took
notice of market needs. GPRS-EDGE is less than an optimum solution and there
are many more effective ways to deliver wireless data, but those other solutions
cannot deliver such a desirable combination of services to such a short timescale.
Look at the marketing opportunity here - the way Microsoft did in the US - it is
massive. Moreover, the market is virtually screaming for packet-switched data.
It does not necessarily want something that is easy for the industry to deliver -
HSCSD. The industry also has to look beyond the current limitations of a
technology and recognise that in order to create a market you have to start
somewhere. End users will appreciate something that is better and cheaper than
the mobile data services on offer now, and if it improves in the future, even
better. In fact, better could well be IMT-2000, along with the chosen air
interface. Start delivering this killer combination of services, the customers will
come, and the momentum behind the technology could be unstoppable. But will
it happen? There are some interesting mobile data developments in niche
markets, but no grand scheme, so far.
When Sonera introduced its IP Communicator service late last year, it was keen
to point out that the global Internet telephony networks of carriers such as
Worldcom and Delta Three only offered toll bypass, that is, there are no
value-added services. However, they could buy their way into GSM and deliver
IP all the way from the desktop or smart mobile phone through to network
servers, offering the killer communications applications this technology can
deliver.
Sonera is bringing some interesting features to the local market, a fact that was
not lost on fixed line operator RSLCom Finland, which recently signed up as
distributor of Sonera's mobile calls. Corporate customers use IP PBXs running
software from Selsius Systems (see Computer Telephony, CI, December
1998) . Using a regular browser they can access a Sonera server that runs
personal profiling applications, fax, voice mail, call forwarding and the like. The
end user device can be a PC, a traditional phone, an IP phone or even a mobile
phone. Seamless interoperability is guaranteed by conformance to standards,
most prominently H.323.
According to Mika Uusitalo, chief technology officer, "For the first time, Internet
multimedia applications as well as intelligent telephony (IN) services can be used
and managed by the user through one Web interface." If Sonera wanted, it could
link the IP Communicator network into one of those global IP networks and
applications hosted in Finland could be used anywhere in the world. Or, one of
the global IP players could buy Sonera. There have been stranger acquisitions.
Sonera has demonstrated what PTOs can and should do, but most of them lack
the vision to be able to do it and, even now, fail to understand the seismic impact
that IP will have on their well-ordered worlds.
As for the GSM operators, they are still obsessed by the numbers game and
signing up more and more subscribers. They seem oblivious to data and the
provision of data services. If they are not careful, even the wireless operators,
still relatively new on the scene, might find themselves as focused on the past as
their PTO fixed-line forebears.
It's up to them, either to go the way of the dodo [So, this is where the short on the yahoo board stole the idea...], or to act now to develop
strategies that will transform cellular and Internet technologies into a powerful
and dynamic single entity. Now, that is a consummation devoutly to be wished.
Ten reasons to go GPRS
1 GPRS in the WAN enables low-cost data communications and you're
always on-line - there's no dial-up.
2 GSM operators are already doing IP conversion, so you get
packet-switching end-to-end.
3 IT sees GPRS as if it were IP, so network management is much easier.
4 GPRS can also carry voice, although the industry tends to keep quiet about
this capability.
5 When converted to IP, GPRS traffic can be carried over the new global
networks. The mobile phone then gets carrier-class voice over IP (VoIP) for
the cost of two local calls.
6 The technology makes extremely efficient use of the available spectrum.
7 The medium could even provide wireless LAN functionality in small offices.
8 GPRS and IP = wireless Internet and intranet.
9 It's proven technology and implementation could start tomorrow.
10 3G networks will be backwards-compatible with GPRS, so it's a
future-proof investment.
GPRS to the rescue
Multi-slot GPRS takes data
rates to around 400kbit/s,
making it a candidate
technology for Internet
access. While cable and
xDSL technologies offer
higher rates, availability of
these services is limited. The
data rate is usually
determined by Internet
bottlenecks.
This scenario relies on GSM
network operators doing
protocol conversion and
working with an ISP. Many
operators do IP conversion for their corporate clients. Packet switching allows
users to be online all the time and tariffs are based on the amount of data
transmitted, not the duration and distance.
(1) A home PC user, (2) a mobile data user and (3) a GSM subscriber
connect by radio waves to (4) the combined ISP/GSM operator operating an
IP Gateway server. At the receiving end is another phone user (6) served by
the corresponding GSM operator (5) operating an IP Gateway. This user is
receiving the call from (3), the Internet being used to make a toll bypass. A
wireline link, possibly ISDN, also links the next generation operator to a
router (7) connected to the host network of an enterprise. ISDN enables
quick set-up time and allows data compression to be used. When the network
operator does IP protocol conversion, mobile workers (8) can communicate
via a router and therefore obtain access in the same way as workers who
work on the LAN.
Voice calls made over GPRS will have to offer lower tariffs than regular GSM
calls. Voice calls made in this way can be converted to IP and transmitted
over one of the new global IP networks. This eliminates international tolls in
the same way as Internet telephony. This scenario is not too far from
realisation: Ericsson is to implement GPRS in Deutsche Telekom¹s nationwide
GSM network. T-Mobil is planning to offer wireless Internet and other
IP-based services of up to 115kbit/s.
C O M M U N I C A T I O N S©
I N T E R N A T I O N A L
|
