Spread spectrum: a protocol with a past and a future.
Spread spectrum is probably today's most important wireless networking protocol. As unlikely as it sounds, a twenty-six-year-old Hollywood movie star, a screen siren named Hedy Lamarr, invented spread spectrum. In 1942, she and a partner (who was a composer and music producer) received a U.S. patent (#2,292,387) for the invention. The essential insight in spread spectrum is to spread a signal over a wider bandwidth than normally would be used, which makes jamming or intercepting that signal much more difficult. This idea has military significance because secure communications on the battlefield are a high priority for any army. During World War II, the U.S. government, realizing this, quickly snapped up the patent and classified it as "top secret."
Ms. Lamarr was raised in Austria and, as a young woman, was married for four years to a much older industrialist who built armaments for Adolf Hitler's war machine. He was jealous and refused to let her out of his sight, thus requiring her to attend many of his business meetings each day. So, she was a quiet observer in the background. Most of these meetings focused on war machinery, and she was therefore well-schooled in such matters. She came to despise the Nazis and her husband. As a result, in the late 1930s, disguised as a maid, she escaped from her hotel room while they were on a business trip. She went to London and then made her way to Hollywood, where she became a major movie star. One afternoon in 1940, while gazing at the Pacific Ocean, she was playing a piano. As her fingers moved from key to key on the piano, she realized that hopping from frequency to frequency while transmitting radio communications on a battlefield would make it much harder to jam the signals. It would make wartime transmissions much more secure. Eventually, her ideas were applied to torpedo guidance systems for submarines, as well. Hedy Lamarr never made a cent from inventing spread spectrum. But today, various forms of spread spectrum are among the most important concepts in wireless communications, and her protocol is now imbedded in numerous wireless products.
Every telecommunications protocol standard has its own language and terminology, and this one is no exception. The concept of spread spectrum is based on a sequence of digits known as a spreading code. At the source, this code is used to spread a signal across a wider range of frequencies; the same code is used at reception to return the signal to its original form. Spreading the signal in this manner uses more bandwidth for transmissions. Therefore, on the surface, this approach would appear to waste a critical resource. But spreading the signal has very significant advantages. Spread spectrum can provide the military with immunity from jamming, yes, but the same characteristics provide commercial networks with valuable immunity from various forms of interference, noise, and distortion. Spread spectrum protocols also can be used for hiding and encrypting signals because a receiver cannot decode the incoming signal without the original spreading code. Most importantly, several users can use the same range of high-bandwidth frequencies at the same time with very little effective operational interference. These properties make spread spectrum especially desirable for cellular telephony applications.
There are three forms of spread spectrum: frequency hopping spread spectrum, direct sequence spread spectrum, and code division multiple access. With frequency hopping spread spectrum (FHSS), the signal is broadcast over a seemingly random series of radio frequencies, hopping rapidly from frequency to frequency at fixed intervals (tiny fractions of a second). A receiver, hopping between frequencies in synchronization with the transmitter, picks up the message. The sequence of channels used is dictated by the spreading code, and both the transmitter and receiver must use the same code to synchronize transmissions. Any would-be eavesdroppers can hear almost nothing, a few blips and pops maybe. Attempts to jam the transmission on one or several frequencies only knock out a few bits of the signal, but for voice transmissions, loss of a few bits here and there is of no consequence in an overall communication. FHSS is the oldest form of spread spectrum and is the one that the movie star invented.
The other two forms of spread spectrum are more sophisticated. Both allow users to share the same bandwidth simultaneously without interfering significantly with one another's signaling. For direct sequence spread spectrum (DSSS), the transmission for each bit in the original bit stream actually is done over multiple channels simultaneously using a different kind of spreading code. This code is a fixed pattern of bits; it is used with an algorithm and one bit from the original bit stream to create a different pattern of bits that is transmitted in parallel, one bit on each channel, to a receiving device that is set up to monitor each of these channels. Any authorized receiver will have the same algorithm and will know the spreading code, so the receiver will be able to reconstruct the original bit from the transmitted spread pattern. The beauty of this system is that even if some of the channels are blocked by interference or by others using those channels at the same time, the receiver still can reconstruct the original bit from the portion of the transmission that succeeded in getting through. DSSS improves network reliability dramatically over previous approaches.
The last, and most important, form of spread spectrum is code division multiple access (CDMA), which is used in cellular telephony. This technique is relatively new and employs and extends the DSSS approach. CDMA systems mix a long binary spreading code called a user code with a small amount of communications data to produce a combined signal that is then spread over a very wide frequency band. The same user code is used at the destination to reconstruct the original digital signal. In this approach, every device that connects to the system (such as a digital mobile cell phone) is dynamically assigned a unique user code when a connection is established. This code is typically more than a hundred digits in length. The number of digits used in the code is called the spreading factor. The code uses binary digits, with the digits being interpreted as plus or minus ones. So, each active device in the system is associated with a unique code made up of a long sequence of plus ones and minus ones.
When a cellular device communicates with a cellular tower, the user code is transmitted across multiple channels to indicate a one, and its complement is transmitted to indicate a zero. (The complement of the user code is the same kind of code with all the pluses and minuses reversed.) Both the device and the tower use the user code and its complement to communicate with one another. The number of channels used for transmission is the same as the spreading factor, which is the number of digits in the user code, so the entire pattern of plus and minus ones arrives at the destination for each transmission, in unison. The receiver then decodes the incoming signal to get back the original bit. A stream of these transmissions effectively sends a bit stream between the cellular device and the tower, as required.
But wait a minute! The cellular airways can be jammed full of traffic. Using spread spectrum means that channels are shared and allocated bandwidth is overlapped. How does a receiver figure out if a message it hears in the air is meant for it, and what is being sent? This is the ingenious part of CDMA. All it requires is a bit of mathematics. The computer in the cellular tower knows the user codes for all active devices in its area because it assigned those codes originally. The codes are just sequences of plus or minus ones; they can be treated as vectors and manipulated using vector algebra. When a transmission is received, the computer can quickly calculate a dot-product between the received code pattern and each of the user codes for all the active devices in its area to answer these questions.
Call the spreading factor "k." Among all of those dot-products, one will be with the user code for the transmitting device. When the computer calculates a dot-product using that user code and the received code pattern, it is really multiplying that user code by itself. It will get either +k or -k. That result cannot happen with any other user code, and that identifies the sender. If the result was +k, then the sender transmitted a one; if the result was -k, then the sender transmitted a zero. So, these dot-products identify who sent the transmission and what it was. Thus, the bandwidth can be shared among users and is used efficiently, which is one of the great benefits of spread spectrum. If the cellular tower's computer is fast enough, it can handle all of the traffic flowing to it using this scheme for many, many users. The computer also can assign user codes so that the other dot-products will result in very small numbers compared to k. These are called orthogonal (or near-orthogonal) vectors.
Another great benefit of using spread spectrum is robustness. For example, if during transmission, interference problems block receipt of some of the minus ones or plus ones that were sent, then the dot-product for the sender's code might be +89 (because of missing data) even if k is (say) 101. But all the other dot-products will be very small numbers by comparison, probably between plus and minus twenty. So, the computer (by selecting the user whose code gives the largest absolute value) still knows who sent the transmission and what bit was sent! The information still gets through, even if part of the signal has been destroyed in route. So, this is a very powerful protocol indeed.
Historically, one of the greatest problems with cellular telephony has been finding a way to share bandwidth efficiently. Failure to share dramatically limits the capacity of a cellular system and severely restricts the number of devices that an individual cell tower can support at any one time. Spread spectrum protocols are changing all of this. And these limitations are lifting. The future of cellular telephony lies with spread spectrum. And to think that it all began with Hedy Lamarr!
Charles K. Davis is a Professor of Management Information Systems at the University of St. Thomas in Houston, Texas. He is an authority on the use of information technology in business. In addition to authoring more than a hundred articles, monographs, books, and reports, he has held numerous technical and managerial positions with IBM Corporation, Chase Manhattan Bank, Occidental Petroleum Corporation, Pullman Incorporated, and Deloitte & Touche. He received a PhD in MIS from the University of Houston, an MBA from Columbia University, an MAT from Harvard University, and a BS from Oklahoma State University. Dr. Davis has also served as a Fulbright Senior Specialist with the Universidad del Valle de Guatemala and is an Information Management Fellow of the AIB Centre for Information and Knowledge Management at the University of Limerick, Ireland.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Forum on Business & Economics|
|Author:||Davis, Charles K.|
|Publication:||Phi Kappa Phi Forum|
|Date:||Sep 22, 2006|
|Previous Article:||Forum on education & Academics.|
|Next Article:||Alfred Woodcock, a natural scientist.|