When many people think of 2.5G, they think of infrastructure: the network
upgrades necessary to support the higher data rates and advanced,
packet-based services. But if that's all they think about, they're missing a
big piece of the 2.5G puzzle. The handset upgrades necessary to support
services such as GSM's GPRS and CDMA's 1XRTT will be just as vital.
In fact, handset sales today represent an even bigger portion of the
wireless business than infrastructure. Goldman Sachs estimates that 1999
handset sales were $50.4 billion, compared with infrastructure's $47.7
billion, and that handset sales will experience triple-digit growth rates in the
next five years.
But the handsets sold five years from now are likely to operate much
differently than those of today. The main differences are in five areas:
increased processing power, connectivity concerns, additional
applications, audio upgrades and display upgrades.
Increased Processing Power
The move from a voice-centric, circuit-switched world to a data-centric,
packet-based world means that phones will require increased processing
power. With packet, rather than moving information over one dedicated
channel, information moves across several different channels. In short, it's
transmitted by squeezing packets of data in unused channel space, so a
link is used only when there's information to send.
This difference affects handsets in two ways: It requires additional
processing power, and the way the chipset processes information must
change, so the internal makeup of the chipset also must change. The move
from a 1-channel to a 2-channel configuration requires that processing
power nearly double in order to transmit data at the same speed, and that in
turn saps battery life.
Several solutions exist. Because packet data is based on bursty
transmissions, the phone can spend more time in a quasi-alert state, which
preserves battery life and offsets some of the battery resources drained by
the higher processing power. Second, innovative chip manufacturers such
as PrairieComm are developing more modular chips with sophisticated
algorithms and using peripherals such as accelerators to increase speed
without having to increase processor size or cost.
The GPRS standard itself also helps. For one, it's dynamic: The same slots
used to send data for one transmission can be used to receive data during
the next transmission. It's also tiered: The specifications allow
manufacturers to build handsets with anywhere from zero to eight slots.
For instance, the Type 1, Class 12 specification allows a total of five slots,
and a maximum of four out of those five slots can be dedicated to transmit
or receive in any one direction. Each slot can support 14.4kb/s of data, so a
total of four slots can transmit slightly more than 56kb/s of data in any one
direction. With Type 2, the functionality becomes even more advanced:
Each slot can support data moving in both directions during the same
transmission, so all eight slots could be used to transmit and receive for a
maximum rate of 115kb/s.
Although the standard allows for maximum efficiency in transferring data,
the first generation of GPRS phones likely will be Type 1, Class 12, which
supports only five total slots that must either transmit or receive but not
both during each transmission. Thus, despite advertised rates of 115kb/s,
the maximum rate in one direction will peak at slightly above 56kb/s, while
the maximum rate in the opposite direction is 14.4kb/s. So, for example, a
handset that's combining four slots to receive at a total rate of 56kb/s
would be able to send data at the same time at only 14.4kb/s because that's
all the remaining slot can accommodate.
CDMA takes a similar approach to maximize efficiency. In 1XRTT, the
transmit and receive functions always are negotiable, so the base station
and the mobile can work together to decide what data rate can be
supported. For 1XRTT, the advertised maximum rate is 144kb/s, which the
first handsets are expected to support.
Connectivity Concerns
Most people assume that when the first 2.5G phones hit the market, many
user applications will reside on other devices, such as PCs. As a result, the
process of transferring the data out of the phone and into an application
will be key.
Bluetooth or cables likely will be the primary mediums for that transfer. It's
vital to integrate connectivity technologies into the chip and even more
important to ensure that the necessary software connections can support
high data-transfer rates.
But not all applications will reside outside the phone. In fact, as phone
memory increases, more applications will be put in phones. One of the
most-talked-about phone-based applications is MP3, which allows users to
download music from the Internet and make the phone double as a portable
stereo. To support MP3, more software must be added to the baseband
chip and into non-volatile memory.
Looking & Sounding Better
Applications such as MP3 will lead to improved audio. Today, vocoder
quality is just good enough to ensure that people don't sound as if they're
speaking into a tin can. Downloading music requires even more improved
audio quality, which in turn demands more processing power.
Video won't be far behind, but today's black-and-white screens, which
operate using just one bit per pixel, simply won't do. Like audio, video
demands more processing power. Color screens that support video could
require as much as 20% more processing power than current
black-and-white displays. Users will expect all of today's features,
improved battery life and new features such as audio and video on par with
consumer-electronics products, all at an affordable price.
Although these new capabilities mean more expensive phones, carriers
must realize that processing power will become even more important for
ensuring that handsets keep pace with consumers' needs. As a result,
carriers must pay attention to which chips are used in their handsets and to
work with chip manufacturers to develop the functionality that users
demand. PrairieComm, for example, already is working with carriers such as
AT&T Wireless to determine and develop the functionality necessary in
future portable devices.
Carriers also should check whether the chip manufacturer develops both
the hardware and the software for its chipsets. Those that develop both
can optimize the integration of the two to produce higher quality links,
longer battery life and faster, more efficient data rates.
Despite 2.5G's many challenges, Moore's Law suggests that technology
should be able to keep pace with the demand for greater processing power.
But of course, as memory increases, there always will be applications to fill
it. As Parkinson's Law of Data predicts, “Data expands to fill the space
available for storage.” In other words, bandwidth will be consumed as fast
as it's made available. For wireless, that means the industry has its work
cut out.
Wireless Review |
