Electronic and Mobile
Commerce: Wireless Networks
The widespread adoption of the Internet in the past decade
was driven in part by the continual emergence of electronic
commerce (e-commerce) applications that took advantage of
the new technology, and, of course, the number of innovative
applications and money-making enterprises in turn grew dramatically
with the rise of the Internet. This self fuelled expansion
reached its zenith (or nadir, depending on which way fortunes
turned) with the boom and bust of many of the newly formed
companies behind these applications. The imagined potential
of the new technology in many cases could not be borne out:
there were too many trying to achieve the same thing at the
same time, but, mostly, the business models behind applications
were simply not good enough. The Internet has provided the
means to carry out every day business transactions, such as
shopping and managing bank accounts, in new ways, and also
inspired new revenue-creating opportunities such as charging
for downloadable video and music content or providing forums
for people to get into contact with each other via email (www.friendsreunited.com,
for example, which charges for divulging the email addresses
of other users). Often, businesses can save a lot of money
by conducting business online, cutting the overhead costs
of employees and office space: several banks and airlines
operate entirely over the Web.
However, the imminent full introduction of the latest wireless
network technology may offer even greater potential for e-commerce
applications. Whereas the majority of Internet based services
primarily differ from their traditional forebears only in
the way they are provided, wireless networks could deliver
location-specific online services such as localised weather
or traffic reports, and local area maps. Moreover, users tend
to have mobile devices near them at most times, and as far
as online services are concerned, they are potential customers
as long as the device is close. Also, since the user of a
mobile device can be found wherever his or her device is,
passive applications such as email delivery, news updates,
or advertisements for local businesses, would come into their
own: users would receive all this information instantly rather
than only when he or she sat at a desktop machine. The Gartner
Group and other research firms predict that, by 2004, the
installed base of mobile phones worldwide will exceed 1 billion
– and this considers only mobile phones: the numbers
of other mobile devices such as PDAs are also expected to
grow dramatically.[1]
It is understandable why the technology industry is putting
so much money behind the development of the infrastructure
needed for mobile e-commerce (m-commerce): the rewards stand
to be enormous. In what follows, we shall briefly survey the
core technologies underpinning m-commerce.
Wireless networks can be roughly divided into three categories:
system interconnection, wireless LANs (local area networks),
and wireless WANs (wide area networks, spanning large geographical
areas). Wireless system interconnection is the interconnection
of the components of computer, such as mouse, keyboard, printer,
scanner, or digital camera, using short-range radio. The short-range
wireless network designed to do this is known as Bluetooth.
Peripheral devices can be connected to a computer simply by
being brought within range: no cables or driver installation
are required.
The desirable simplicity of this option is obvious. In its
simplest form, system interconnection networks use the master-slave
paradigm where the computer is the master and the peripheral
devices slaves. The computer tells the slaves what addresses
to use, what frequencies to use, when and for how long to
transmit, and so on. But wireless devices are not necessarily
mobile devices. Fixed machines can be connected to networks
wirelessly – this is often advantageous in situations
where laying cables is costly or impractical – and mobile
devices such as laptops may have to be connected by cable
to a network to access it. Bluetooth does not really have
much impact on m-commerce and, though we include it merely
for completeness, we will not consider it further.
Wireless LANs are networks in which each machine has a radio
modem and antenna with which it can communicate either with
a base station (such as a central antenna) or with each other.
Most systems implement the IEEE 802.11 (or Wi-Fi) standard
for wireless LANs. Wireless LANs provide mobility in so far
as a device can be connected to a network without having to
actually plug it in. Wireless communication only takes place
over short distances and the majority of data sent and received
by a device in a wireless LAN will have travelled over a line
based network – probably the Internet – for nearly
all of its journey. However, wireless LANs offer the possibility
of mobility because devices can roam between LANs, thus essentially
maintaining their connection to the encompassing network –
again probably the Internet – while on the move.
Mobile phones use the third type of wireless network, wireless
WANs. In certain respects wireless WANs are like wireless
LANs, but the distances between communicating devices and
base stations are much greater – measured in kilometres
rather than metres – and the bit rates much lower: wireless
LANs can operate at rates up to about 50Mbps over distances
of tens of metres, while bit rate across wireless WANs is
currently below 1 Mbps. There have been three generations
of wireless WANs – or mobile phone networks –
and there are high-speed wireless WANs being developed that
could bypass the telephone system providing high-speed access
to the Internet to homes and businesses. The standard behind
the development of this technology is known as IEEE 802.16.
We shall discuss wireless LANs later when we consider them
in relation to wireless WANs, but since mobile telephony is
the prevailing mobile technology, we shall consider it more
fully.
Although no longer state of the art, certain properties of
first generation analogue mobile phone networks have been
directly inherited by its digital successors in order achieve
backward compatibility. In all mobile phone systems, a geographical
region is split up into regions (or cells, and hence ‘cell
phone’). Each cell has its own radio frequency band
allocated to it, and though frequency bands are reused across
the network as a whole, no two neighbouring cells share the
same frequency, thus avoiding interference. All phones transmit
to the base station in the middle of the cell they occupy
at the time. At any instant a phone is logically in one specific
cell and under the control of that cell’s base station.
Base stations are in turn connected to a Mobile Telephone
Switching Office (or MTSO), the ‘nerve centre’
of the system. MTSOs communicate with the base stations, each
other, and the rest of the network. When a phone leaves a
cell, that cell’s base station senses the fading signal
of the phone and asks the base stations of its immediate neighbours
how strong a signal they are receiving from the phone. The
original base station hands control of the phone over to the
new base station that is receiving the strongest signal, which
should be the cell in which the phone is now located. The
new base station informs the phone of the handover and, if
the transfer is taking place during a call, the phone is told
to switch to a channel from the set of frequencies used by
the new cell, since the frequencies of the old cell are of
course not used by any of its adjacent cells. This whole exchange
is known as ‘handoff’, and the assignment of the
new channel is handled by the MTSO.
There are two types of handoff: soft and hard. Soft handoff
is where the old base station does not relinquish control
until a new one has taken over. This enables a seamless transition
that is not noticed by the user, but it requires the mobile
device to handle two frequencies at the same time. But current,
first and second generation (2G), networks work with hard
handoff. This is when a mobile device is dropped before it
is picked up again. There can be a break in the signal noticeable
by users but it is inevitable with the current design. If
there is no new base station to take over, the signal will
simply be lost. 2G networks do, however, use what is known
as mobile assisted handoff where the device itself requests
a new connection from its new base station.
Mobile phones have unique identifiers that allow them to
be picked out when sending and receiving calls and SMS messages.
A phone has a home MTSO that always knows what cell the phone
is located in. When making a call a phone transmits the number
of the phone it wants to call together with its own identity
to the base station, which forwards these to the MTSO. The
MTSO locates the phone being called, looks for an idle radio
channel, and having assigned one, sends the channel number
to the calling phone, which waits for the other phone to ring
and be answered. All transmissions to phones go via their
home MTSO to find out where they are. All idle phones listen
for messages directed to them. When a message from the base
station of the cell the phone is occupying – essentially
of the form “ Are you there?” – is received,
the phone responds and a connection is established with the
calling phone. Much detail has been glossed over, but this
should serve as a general picture of the process underlying
communications between mobile devices on wireless WANs. From
the perspective of m-commerce, it is important that the location
of mobile devices, and therefore their users, is always known.
A forerunner of the high-speed wireless WANs mentioned above
is the third generation (3G) of mobile system. Among the factors
driving the mobile industry, a primary one is the exponential
growth of data traffic over fixed networks: data traffic now
far exceeds voice transmissions. The assumption is that, as
long as the technology is developed to support it, mobile
networks will match this pattern. Secondly, since most media
has become available digitally, there has been a rush to converge
the telecommunications, computer, and entertainment industries.
The goal of the backers of 3G mobile networks is to have small,
mobile devices that do duty as a phone, music and video player,
game machine, email client, Web interface, and more, with
wireless connectivity to the Internet at high bandwidth. The
potential of 3G for m-commerce applications is clear. With
internet access available to a user at all times, exposure
to online services is accordingly increased. Furthermore,
these services are instantly available worldwide with automatic
connection via a satellite when no terrestrial network can
be located, and with quality-of-service guarantees.
However, there are problems with 3G. Backers of the technology
have paid vast sums of money in both Europe and the US at
auctions for licenses to use the necessary radio bandwidths.
This has made companies particularly desperate to make the
money back, and in some cases this seems increasingly unlikely.
3G has long been the dream of would be m-commerce businesses,
but it is one that has been a long time coming. As well as
the high costs involved the technology has been harder to
implement than anticipated and there have been devastating
delays with the development of suitable mobile devices. As
Lee Garber notes, “the advantage of 3G will be its ability
to support a wide variety of different applications, and that
will be the killer characteristic of 3G”[2]
. But 3G technology has taken so long to bring to market,
that some cynics believe the hype has died and people will
pass it by. Users may decide to get wireless services from
other types of technology. So many 802.11 wireless LAN access
points are being installed all over the place that people
may just be able to wander from one 802.11 access point to
another rendering 3G networks obsolete. “Although 3G
is not designed to compete with WLAN [wireless LAN], it will
have to do so” [3],
notes Garber again. Many wireless LANs are already deployed
and so have a head start on the market, users may become accustomed
to wireless LANs and be sceptical of what 3G systems offer
over the existing one. Accepted technologies are often unwillingly
given up. In fairness though, wireless LAN systems would need
to provide more mobility-oriented applications to avoid simply
providing wireless versions of desktop software. Wireless
LANs’ principal advantage over 3G is the much lower
cost of development, but the technology is not suited to wide-area
coverage. 3G’s ability to locate users geographically
also offers the potential for companies to provide many attractive
services that would be unavailable through 802.11. Arguments
swing both ways; it is still too early to know the fate of
those who have invested in 3G technology.
Just as TCP/IP protocols, and general purpose Web browsers
were principle drivers of Internet growth, wireless services
are sensitive to a similar need for consensus across countries
and companies in the technology adopted. After a period of
simultaneous development of several competing standards, agreement
was reached with the Wireless Application Protocol (WAP) governing
the presenting and delivering of wireless services to mobile
devices – the technological layer above the network.
WAP is supposed to supply the WWW to wireless devices with
small screens, slow CPUs, and little memory. WAP does not
use HTML (Hyper-Text Markup Language), but instead makes use
of a markup language called WML (Wireless Markup Language),
which means in principle that a WAP device can only really
view pages on the Web that have been converted into WML. Because
this is obviously a major drawback, an on-the-fly HTML to
WML converter is provided in the WAP architecture protocol
to increase the set of pages available. But still, no images
and only a small amount of text can be viewed on a WAP phone,
and developers of m-commerce sites have to bear this in mind.
Because of the lack of content, WAP1.0 is generally perceived
as a flop. Users were billed by the minute for slow access
to a poor supply of Web pages. Now WAP2.0 has been developed,
and it has adopted many lessons learned from i-mode.
In dramatic contrast to WAP, i-mode, developed by NTT DoCoMo
in Japan and WAPs main competition, had over 35 million subscribers
(who had access to over 40000 special i-mode websites) within
three years of its launch. Unlike WAP, there is no billing
for connection time. Like an ADSL or cable service, the connection
is always on. Users instead pay by amount of data delivered
to the phone, and the most popular services are email and
games. Users in the West are used to paying a fixed monthly
fee for connection to the Internet irrespective of their usage.
For example, Yahoo, Google, or BBC News do not charge for
services and those who use the Internet more than others pay
the same set fee. But in Japan, where fixed line telecommunications
charges are prohibitively high for most, not many people have
access to the Internet from a PC. The business model is therefore
very different. For most i-mode users, the mobile device is
their only means of connecting to the Web and billing is based
not on a flat rate but on services subscribed to and amount
of data delivered to the device. Most official services are
cheap and the billing is handled centrally by NTT DoCoMo (which
keeps about 9% of each service provider’s fee), which
sends users a single monthly bill for all services used.
There is a very bright future for m-commerce, but businesses
must tailor their services accordingly, and develop innovative
ways of making money. I-mode is an example of a successful
combination of technology and content tailored to i-mode’s
typical users – Japanese teenagers and women in their
twenties. This model probably will not migrate directly to
the West, where different fashions and user-habits require
novel targeting, but much inspiration of a more general nature
should be gleaned all the same. Charging methods must also
suit the market. Internet users do not expect to have to pay
for many online services, whereas mobile phone users do. Businesses
hope users of mobile devices – now much more than just
phones – will follow the second model. An explosion
of applications might lead to innovative payment techniques,
such as free services in return for accepting downloaded advertising
perhaps. Indeed, it is likely that advertising will form a
significant percentage of content delivered to mobile devices.
Smart messages could take advantage of being able to reach
individuals at anywhere, perhaps sensing when they are near
the products of services advertised. Or frequent purchasers
of services such as theatre tickets or flights can be reached
instantly whenever special offers or newly available seats
become available. Shops could send coupons and details of
special offers to people as they passed. Indeed, mobile devices
may even replace credit cards, and paying for an item may
soon be no more difficult than tapping in a PIN and waving
the device over an in-store sensor. Applications and wireless
devices will evolve together – applications will not
be developed unless there are sufficient devices to make the
development worthwhile, but these devices will not be built
unless there are enough attractive applications available
to make users want to buy them.
- James Senn: “The Emergence
of M-Commerce”, Computer vol 33, December 2000, p148.[Return]
- Lee Garber: “Will 3G Really
Be the Next Big Wireless Technology?”, Computer vol
35, January 2002, page 31.[Return]
- ibid, page 32.[Return]
- BIBLIOGRAPHY
- Lee Garber: “Will 3G Really Be the Next
Big Wireless Technology?”, Computer vol 35, January
2002
- James Senn: “The Emergence of M-Commerce”,
Computer vol 33, December 2000
- The Economist: “Move Over 3G: Here Comes
4G”, http://www.economist.com , May 29th 2003
|
