with permission. Copyright April 1998 Scientific American Inc. All
Dicing information into digital bundles
and transmitting them at low power over different frequencies can enable
millions of people to send and receive simultaneously.
by David R. Hughes and Dewayne
Conventional wisdom holds that radio airspace
is a valuable -- and limited -- resource that has to be rationed, like
water in a desert. That mind-set comes from traditional transmitters and
receivers, whose operation must be restricted to narrow, dedicated slices
of the electro magnetic spectrum to minimize interference. Thus, governments
have parceled out and licensed radio channels like real estate. In the
U.S., the Federal Communications Commission (FCC) has sometimes used a
cash-bidding process to allocate precious frequency bands for a variety
of purposes, including commercial television and radio broadcasts; military,
marine and police transmissions; taxi dispatchers; CB communications; and
ham radio operators and cell phone consumers.
Recent advances in digital communications,
though, have opened the door to an entirely new model. Transmitters can
now deploy so-called spread spectrum techniques to share channels without
running afoul of one another. Information can be diced into tiny electronic
bundles of 1s and Os and then transmitted over radio waves, with each packet
sent over different channels, or frequencies, at low power. New studies
have shown that, theoretically, millions of radio transmitters within the
same metropolitan area can successfully operate in the same frequency-
band while transferring hundreds of megabits of data per second.
Such shared use of the spectrum challenges
customary practices. In the past, when allocating narrow frequency bands
for exclusive commercial purposes, the government has granted licenses
to firms, such as those offering cell phone and personal communications
services (PCS); Those licensees charge consumers for services, just as
conventional telephone companies bill their customers. In the new economic
model, the middleman becomes unnecessary: consumers can communicate with
one another directly and at no charge, even when they are kilometers apart
and when myriad other people are using the same radio channels. This fundamental
change has led to revisions in the regulatory policies of the U.S. and
foreign governments, which have now designated certain frequency bands
for free, unlicensed use of spread-spectrum radios by the public.
Enough Spectrum for Everyone
What is the basis of this technology revolution?
Traditional radios work by broadcasting information over a single, narrow
channel with high power. By operating in as skinny a sliver of the electromagnetic
spectrum as possible, a transmitter thus makes room for other devices to
operate in neighboring frequencies without interference. But radio engineers
now know that the opposite way of transmitting data -- by smearing,
or spreading, information across a wider chunk of the spectrum at low power
-- is more efficient.
The concept is counterintuitive: instead
of carving a pie into a finite number of small slices, the technology allows
a larger sharing of the whole pie by permitting an all but unlimited number
of individuals to take barely noticeable bites. Although spread-spectrum
radios use more bandwidth than necessary, by doing so the devices avoid
interference because the transmissions are at such minimal power, with
only spurts of data at any one frequency. In fact, the emitted signals
are so weak that they might be almost imperceptible above background noise.
This feature results in an added benefit of spread spectrum: other radios
have great difficulty eavesdropping on the transmission. In practice, only
the intended receiver may even know that a transmission is taking place.
The covert nature of spread spectrum was
initially its main attraction. During World War II, the U.S. military became
interested in an intriguing device that actress Hedy
Lamarr had co-patented, The concept was simple enough -- instead of
broadcasting information over a single channel, where the enemy might stumble
upon the transmission, the device switched channels continually, broadcasting
a little bit of information here and a little bit there, in accordance
with a secret code known only to the transmitter and the intended receiver.
This repeated hopping of frequencies would make it extremely difficult
for the enemy to pluck the entire transmission from the surrounding noise.
But Lamarr's device was deemed impractical because it relied on an unwieldy
mechanical contraption to perform the frequency hopping.
Subsequent advances in electronic circuitry,
however, have made spread spectrum feasible. Semiconductor chips, crammed
with thousands of transistors, can broadcast digitized packets of data
in a seemingly random pattern over many channels. The receiver, designed
to hear the signals in accordance with the precise and proprietary sequence
of the sending radio, is able to pull the fragmented information in the
right order from the different frequencies. In addition, when the receiver
encounters missing or corrupted packets, it can signal the transmitting
radio to resend those packets. Also, a technique called forward error
correction can be used to improve the chances that the data are received
correctly the first time.
Electronic technologies have enabled another
method of spectrum spreading: direct sequence, in which the transmitted
information is mixed with a coded signal that, to an outside listener,
sounds like noise. In this alternative to frequency hopping, each bit of
data is sent at several different frequencies simultaneously, with both
the transmitter and receiver synchronized, of course, to the same coded
More recently, further advances in chip
technology have produced digital signal processors that can crunch data
at high speed, use little power, and are relatively inexpensive. The improved
hardware allows more sophisticated spread-spectrum techniques, including
hybrid ones that leverage the best features of frequency hopping and direct
sequence, as well as other ways to code data. The new methods are particularly
resistant to jamming, noise and multipath -- a frequency-dependent effect
in which a signal reflects off buildings, the earth and different atmospheric
layers, introducing delays in the transmission that can confuse the receiver.
In 1985 the FCC finally permitted the unlicensed
use of spread-spectrum radios by the public, albeit with several restrictions.
The radios must operate under FCC regulations in what are called the unlicensed
industrial, scientific, medical (ISM) bands. More important, the
radios are forbidden to operate at greater than one watt, and the transmissions
must be spread a minimum amount across the assigned spectrum.
These restraints notwithstanding, the 1985
(and later) FCC rules have already spawned the development, manufacture
and marketing of a wide range of "no license required" products. Because
mass manufacturing has yet to occur, spread spectrum products for data
transmission from the 60 or so current vendors carry premium price tags
that have limited the technology mainly to large organizations, such as
businesses, schools and libraries. Today a radio that can handle near-Ethernet
traffic (10 megabits per second, suitable for high-speed computer communications}
up to a distance of about 40 kilometers (25 miles) costs $11,000. Devices
with lower capability -- operation at T1 speeds (1.5 megabits per second)
to a range of 25 kilometers or so -- cost $1,500. For very short ranges,
such as for communications within a building, wireless local-area network
(LAN) cards for personal computers are priced as low as $250.
There is every reason to believe these
prices will drop as manufacturing volumes increase to meet the growing
market demand for higher bandwidth and secure wireless connections from
PCs to the Internet. In the future, people may, for example, routinely
rely on wireless transmission to reach a central system that would then
connect to a traditional network of ground-based lines. We predict
that reliable, secure data radios operating at T1 or higher speeds to a
range of more than 30 kilometers will soon cost less than $500 each.
Where is this technology headed? Transmitters
and receivers are becoming fully digital. That trend, combined with the
rapid development of cell-based wireless systems, will open up a wide range
of services based on spread spectrum. The services will rely on intelligent
networks containing "smart" transceivers and switches as basic components.
Such devices would, for instance, know which of the varied spread-spectrum
techniques to use in a given situation to ensure that the information will
be transmitted without error. These new networks will increasingly involve
a diverse mixture of links and switches, some ground-based, that are owned
by different entities.
The Internet today represents the best
example of the self-regulating mechanism that will be necessary in the
new radio environment. The creation of a similar, decentralized structure
for the optimal sharing of the radio spectrum will require a substantial
effort by a combination of telecommunications experts and entrepreneurs
working with the various regulatory bodies around the world. We believe
the deployment and growth of such a system is achievable through increasingly
"smart" electronics, and we envision a self-governing set of protocols
that are built into these intelligent devices. As advanced radios are deployed;
society must tackle the crucial issue of incorporating both positive and
negative incentives within the network infrastructure itself to make the
best use of a shared common resource -- the radio spectrum.
DAVID R. HUGHES and DEWAYNE HENDRICKS
recently helped to install spread-spectrum radios in Ulaanbaatar to connect
eight scientific and educational institutions in that Mongolian city. Hughes
is a principal investigator for the National Science Foundation Wireless
Field Tests. Hendricks is CEO of Warp Speed Imagineering in Fremont, Calif.
SPREAD SPECTRUM COMMUNICATIONS HANDBOOK.
Second revised edition. M. K. Simon, J. K. Omura, R. A. Scholtz and B.
K. Levitt. McGraw-Hill, 1994.
A CHANNEL ACCESS SCHEME FOR LARGE DENSE
PACKET RADIO NETWORKS in Proceedings of ACM SIGCOMM '96. Tim J. Shepard.
The following Web sites contain helpful
tutorials and other background information: