In my last column, I talked a bit about 3G, and especially the UMTS alternative. Although it would be great to have a single, worldwide standard for 3G, differing engineering philosophies and (mostly) politics have kept this admirable objective from being realized. Interestingly, though, the other major implementation of 3G, CDMA2000, isn't all that different from UMTS. Both, after all, are based on Code-Division Multiple-Access (CDMA), which is a fascinating if somewhat counterintuitive technology and certainly the most important in cellular today.
CDMA is an implementation of Direct-Sequence Spread-Spectrum (DSSS) and radio technology that originally achieved prominence in military communications systems, and then in early wireless LANs. The idea in DSSS is simple -- instead of sending 0s and 1s over the air directly, we convert each 0 and 1 to a longer string of bits, which is the "code." This may appear to waste bandwidth, but the technique in fact improves reliability because damage to one or two bits during transmission need not require that the entire packet of data be resent. Rather, when the signal is de-spread by the receiver, we can use statistical techniques to guess what the original bit was. Using the right codes, we can often guess correctly (and we still use error-checking codes at the end of each packet, regardless).
Now, suppose we pick the codes that are orthogonal to one another, meaning that two properly designed orthogonal codes can actually exist in the same spectrum at the same time and not -- really! -- interfere with each other. We'd give one code to one user and another to a second user and so on, and then, assuming everyone transmits at the same power level relative to one another so that no one station drowns out the others – voila, the multiple-access technique CDMA.
CDMA2000 (you can, by the way, find a lot of information on this subject at http://www.cdg.org/), while basically the same in concept as UMTS, has one very important difference from UMTS -- it is designed to operate in a 1.25 MHz channel, the same as is used in the 2G CDMA standard, which is known as IS-95. This means that carriers need not rearrange their spectrum as the GSM community has had to do, and the rollout of CDMA is thus quite nondisruptive to the carriers' existing operations. There is, by the way, a 5 MHz version of CDMA2000, but it seems unlikely that this technology will be available anytime soon.
That's because, though more spectrum is always better in terms of capacity and throughput (the key drivers behind 3G to begin with), CDMA2000 operators (primarily Sprint and Verizon Wireless in the U.S.) are doing just fine with what's available today. The key CDMA2000 technology is called 1XRTT, and it's a replacement for IS-95, featuring improved spectral efficiency and higher data rates (40-60 Kbps is typical). But CDMA2000 also includes technology called 1xEV-DO, which stands for 1x Evolution, Data-Optimized (or Data-Only). The "x," by the way, is the number of individual radio carriers in the channel; the 5 MHz version of CDMA is called 3XRTT.
EV-DO (often just called "dee-oh") is very important because it allows downlink data rates of (in theory) up to 2.4 Mbps in the current Release 0. A new version of DO, Rev A, will be available next year, boosting downlink throughput to 3.1 Mbps and uplink speeds up to 1.8 Mbps. As always, realizable throughput will vary -- I assume you'll see no more than a third of these numbers, and usually a lot less. The carriers can't and won't guarantee throughput because they have no control over prevailing radio conditions, user traffic types and loads, network latency, and the many other factors that ultimately determine effective data rates.
There is a noticeable gap between HSDPA's peak of 14.4 Mbps and the 3.1 Mbps of EV-DO Rev A, and WiMAX's peak of about 75 Mbps. Not to be outdone, the CDMA2000 community has announced EV-DO (the "1x" is now gone) Rev B, which promises 73.5 Mbps peak downlink speeds via the aggregation of 15 1.25 MHz channels into a 20 MHz Rev B channel -- the same amount of spectrum WiMAX uses, and with similar results. A pre-emptive strike against WiMAX, or a service we'll probably have in a few years? It's too early to tell. Nevertheless, these numbers are amazing considering the 2.4 Kbps barely available when I started working wireless just 15 years ago.
About the author: Craig Mathias is a principal with Farpoint Group, an advisory firm based in Ashland, Mass., specializing in wireless networking and mobile computing. The firm works with manufacturers, enterprises, carriers, government, and the financial community on all aspects of wireless and mobile. He can be reached at craig@farpointgroup.com.
