Home' Technology Review : October 2005 Contents 78
To see how well MIMO actually works, I bought a Belkin
router that uses the Airgo chipset. The router is equipped with
three radios and three antennas, which are intended to increase
its range, reliability, and speed. In fact, Belkin claims an 800 per-
cent improvement in coverage area over the best of current Wi-Fi
routers and a 600 percent improvement in speed.
I had not previously set up a wireless router. And since my
technical skills mostly involve breaking things in ways their mak-
ers had not imagined, I expected a challenge. I popped the instal-
lation CD into my notebook's drive and followed the instructions.
When it came time, I plugged in the router, which boasted four
Ethernet ports to connect to PCs, one port for the modem, and a
plug for a power cord on one end and three nubby antennas stick-
ing out of the other. Windows XP Home Edition, which I use,
protested that I was installing an unknown device that could de-
stabilize my system. Nevertheless, I pressed on.
To get Inter net access, I ran a cable from my DSL modem to
the router, which sits in my basement o ce, and then another
cable from the router to my notebook computer. That
worked ne. But when I installed the MIMO notebook
card and pulled the cable, I lost my connection. It took a
bit of poking around in the software that comes with
the system to realize I needed to tell it to activate. (Okay,
so the tr uth is, I was on the phone inter viewing someone
for this story and complained about the lost connection.
He asked me whether the radio was on. One right-click
of the mouse later, it was.)
Including the basement, our house is about 3,000
square feet, but even in the farthest point from the
MIMO router, I have no connectivity problems and no
drop-o in download speeds. In fact, Belkin claims
good connectivity over 250,000 square feet (the fastest current
version of Wi-Fi, 802.11g, generally covers less than one- fth
that), so I could probably read e-mail while mowing my lawn,
if I felt like it. And the MIMO notebook card works ne in the
local co ee shop, using the existing Wi-Fi standard.
In short, MIMO looks to be a good thing for people and busi-
nesses whose connectivity is spotty and unreliable or who expe-
rience slow data transmission for PCs at the extreme ends of their
current Wi-Fi networks' reach. That's nice, but the importance
of MIMO goes far beyond improving the performance of Wi-Fi.
It is, in fact, "the most signi cant radio technology ever," claims
wireless consultant Craig Mathias, principal at the Farpoint
Group in Ashland, MA. The reason for such excitement is that
MIMO appears to be an excellent answer to a wireless problem
that's existed since Marconi.
"In wireless communication, the biggest problem has always
been multipath interference---signals taking di erent paths and
canceling each other out or blurring the signal," explains Ira
Brodsky, president of Datacomm Research in St. Louis. The re-
sult of this interference is familiar to anyone who's ever had
trouble tuning in a station over a car radio: signal drops---the very
thing wireless networks must be engineered to prevent.
MIMO is designed to turn this inherent problem with wire-
less into an advantage. In a sense, MIMO is to wireless what
multiprocessing is to computing---a way to move data faster by
sending it through multiple channels. Each of the radios on the
MIMO chipset pulls in a signal, and all the signals are then r un
through digital signal-processing algorithms and re-formed into
a single transmission. This use of redundant multiple signals
allows MIMO to increase the reliability and range of transmis-
sions. Indeed, the strategy is widely applicable, and over the next
several years, MIMO technology may make its way into cellular
networks and products that might bene t from wireless trans-
mission, like camcorders and televisions.
The exact origins of today's MIMO technology are disputed,
which seems tting, since there is no agreement on its exact
de nition, either. There are various versions of the technology
with such esoteric names as "spatial diversity multiplexing,"
"beamforming," "antenna diversity," and "channel beaming."
What is clear is that Airgo, headed by CEO Greg Raleigh, who
as a graduate student at Stanford University in the 1990s helped
pioneer MIMO, is the rst company to come up with a version
cheap enough to make it onto the consumer market.
Adding to the confusion around MIMO is that it is to be
the basis for the IEEE's 802.11n standard,
which is not yet a standard and is not ex-
pected to become one until at least 2007.
The bottom line for consumers? People
who buy MIMO products before 2007 will
eventually have to buy new products to get
the full speed advantages of the technique
when the standard has nally been n-
ished, predicts Robert W. Heath, an assis-
tant professor in electrical and computer
engineering at the University of Texas at
Austin. Still, if MIMO keeps improving, it
might be well worth it.
A wireless network's perfor mance isn't based on just one fac-
tor, of course: my router can't download data from the Web any
faster than my DSL connection allows. Network speed is mea-
sured in megabits, or millions of bits, per second, and my DSL
on a good day hits two megabits a second. My router's claimed
top speed is 108 megabits a second. So though I knew the MIMO
router wouldn't let my laptop access the Internet any faster, I did
hope that I could use it to enhance my house's feeble wireless
capabilities. Having noticed new televisions with built-in Wi-Fi,
I imagined using MIMO to zip downloads from digital video
cameras to the TV, or even to take stu from the digital video re-
corder and zap it to one of our other TVs. In fact, I was letting my
imagination run amok, war ns Mathias. MIMO isn't yet fast
enough to handle large amounts of video.
And MIMO won't be a universal solution. In some kinds of
wireless networks, such as sensor or radio frequency ID net-
works, single radios will always be adequate---and cheaper. But
for anything that involves data or voice, MIMO is likely to be
adopted. For those uses, speed of data transfer is important. And
MIMO looks like the technology that will nally let wireless
networks start to close the gap between their speed and relia-
bility and those of wired networks.
Michael Fitzgerald is a freelance writer who lives outside Boston.
A frequent contributor to Technology Review, he has also written
for the Economist and the New York Times.
In a sense,
MIMO is to
is to computing---
a way to move
data much faster
Links Archive December 2005 January 2006 September 2005 Navigation Previous Page Next Page