UNFIXING FIXED WIRELESS

MARCH 13TH 2003

The Economist

 

New wireless technologies for delivering broadband to the home are

about to challenge DSL and cable

 

WIRELESS means mobile--or so many believe. Whether it is the fuss over

third-generation (3G) mobile phones and multimedia on the move, or the

unexpected success of Wi-Fi hotspots that give laptop users fast

internet access without having to plug into a telephone socket, people

think of "wireless services" as something to use while away from the

home or the office. But another type of wireless technology may be even

more important. Known as "wireless local loop" (WLL), it could reignite

the consumer broadband market.

 

 The telecoms boom left a glut of fibre-optic cable along trunk routes,

but this links directly only to the largest customers. Homes and small

offices that want high-speed internet access usually subscribe to

either a digital subscriber line (DSL) or a cable-television service.

Both are far from ideal: the phone wires used by DSL and the television

cables tend to be owned by monopolies, and neither was designed for

surfing the web. Retrofitting a 1950s telephone line for broadband

takes a lot of work, making cheap DSL hard to supply profitably.

 

 In principle, WLL has no such drawbacks. Indeed, many see it as an

ideal solution to the local access problem. Radio waves already reach

everybody's home or office. So there is no need to dig up streets or

shoehorn data into a system designed for voice. Of course, most WLL

systems require their own dedicated radio frequencies, but regulators

have been fairly generous with these--selling enough licences to

competing WLL operators at a fraction of the prices paid by

mobile-phone operators. Some can even use the same free, unlicensed

frequencies in the 2.4 and 5 gigahertz bands as Wi-Fi, opening up the

market to anyone.

 

 At least, that is what WLL proponents say. In the real world, wireless

has so far lagged behind both cable and DSL. The problem is that, until

now, most WLL systems were fixed and required a clear line of sight

from the customer's building to the service provider's base-station.

Like satellite dishes, they had to be aligned by expert technicians,

and could be knocked off target by snowstorms or wind. Unlike satellite

dishes, their reception was easily disrupted by events on the ground.

People who had a system installed in winter often found that it stopped

working in spring, as trees sprouted leaves that blocked the terminal's

view of the base-station.

 

 The latest WLL technology could change this. Unlike its predecessors,

it does not require a line of sight, or outdoor installation. San

Francisco-based Soma Networks, for example, has developed a terminal

that can be placed almost anywhere in a customer's building. Like a

mobile phone, it can even be carried around and used from other points

within radio range of a service provider's base-station. The terminal

is little bigger than a video cassette but has to be plugged into an

electric outlet. This is not particularly convenient, but portability

is not its main selling-point.

 

 Rather than the ability to move around, service providers are

attracted to the new WLL's "self-provisioning" feature. Users can

simply buy Soma's terminal in a store and plug it into their PC and

telephone as soon as they get home, without needing to call out a

technician. This is cheaper for the provider, and more convenient for

the customer. The first time the terminal is switched on, it directs

web surfers to a portal where they enter their billing information and

choose which voice or data package they want.

 

 The device is capable of transmitting or receiving at up to 12

megabits per second--about ten times faster than most DSL connections

and more than 200 times faster than a dial-up system. This way,

customers can upgrade to a faster service at the click of a mouse. As

more customers sign up, service providers simply add more base-stations

because all can use the same frequency without causing interference.

Previous WLL systems required adjacent cells to use different

frequencies, making it hard to increase network capacity with demand.

 

 Exactly what service packages will be on offer is up to the network

operator, but most WLL manufacturers envisage pricing based on service

guarantees and peak speed, not volume of data or time spent online. The

network is able to prioritise traffic, ensuring that guarantees are

honoured during busy periods. Even more importantly, it means that

lucrative voice traffic--the service from which telephone companies

still make most of their money--can be given priority over data. Rather

like text messaging on a mobile phone, data travel over spare network

capacity not used by telephone calls.

 

 The similarity to mobile phones is not coincidental. Soma based its

hardware around UMTS (Universal Mobile Telecommunications System), the

3G technology that is slowly becoming available in Europe and Japan.

Entering a market dominated by the likes of Nokia and Ericsson may seem

ambitious, particularly when that market's size and even existence are

doubted by many analysts. But Soma is not really trying to beat the

mobile-phone firms at their own game. The company has left out the

parts of UMTS that deal with "hand-off", the complex process that lets

users move between cells without dropping calls, and it does not have

to worry about miniaturisation or battery life.

 

 Other companies have had the same idea, although most have focused on

different applications. Another Silicon Valley start-up, IPWireless,

has also cherry-picked the UMTS standard, while Texas-based Navini

Networks has its own closely related technology. Both of these are

aiming for true mobility, retaining the hand-off circuitry and trying

to make terminals as small as possible. (They are currently about the

size of a videotape, but both firms claim that cards which can fit

inside a laptop are coming soon.) In the process, however, they have

sacrificed telephony and focused solely on data-only systems.

 

 Not every WLL company is adapting 3G technology. Flarion Technologies,

a start-up firm backed by Cisco Systems, among others, has an approach

based on OFDM (orthogonal frequency division multiplexing), the

technology used in the fastest types of Wi-Fi and some

digital-television systems. OFDM can reach even higher data speeds than

3G, by splitting a high-speed signal into several lower-speed

transmissions and sending each via a different path. Flarion's system

should provide a similar user experience to Wi-Fi, but covering a whole

city rather than only a few hotspots. Like IPWireless and Navini,

Flarion lacks genuine telephony support, gambling that internet

services will be more important in the long term.

 

 They could be right. It is already possible to make phone calls

through all three systems using VOIP (voice over internet protocol), a

technique that carries calls across the internet or other networks

based on the same architecture. VOIP requires special phones, and most

people think it still sounds inferior to normal telephony, but it is

improving all the time. Flarion hopes that its technology will

eventually be embedded inside VOIP cell phones, though for the moment

it is concentrating on cards for laptops.

 

 How long it will be before you can use this technology depends on

where you live. America's three big long-distance carriers--AT&T,

Sprint and WorldCom--all have WLL spectrum licences. To date, only

WorldCom has selected a system, but its roll-out has been delayed by

its bankruptcy filing. Soma, IPWireless, Navini and Flarion have also

shipped systems to smaller service providers in America and to larger

carriers in Japan and South Korea. All agree