Ask any driving enthusiast-and I would be one myself if I didn't live in Boston-what the single most important improvement to his or her car might be, and I'll bet you the answer is new tires. I discovered this long ago, in the prekids years, when I replaced the stock tires on my sports car with a suitably high-performance set. Night and day.

The reason tires have such an impact on the auto is that they're the only part of the rather sophisticated mechanism that actually touches the medium on which the machine operates. The greatest engine in the world is just a hunk of iron (OK, aluminum) if the vehicle it powers doesn't hug the road.

Which brings me to antennas. Granted, the innovations in radio design in recent years have been amazing. CDMA, OFDM, rake receivers, QAM-you name it. But all would be useless without the ability to get the signal on the air and off again. Antennas-the only part of the vehicle that touches the medium.

Years ago, antenna meant dipole. (I'll bet at this point you're wishing you'd paid more attention in that RF propagation class.) And for many years, simple diversity was more than sufficient. Data rates were low, power levels were high and range was typically rather short.

Today, of course, we expect wireless LANs (WLANs), for example, to operate at very high data rates indeed, both indoors and out, and both cellular and terrestrial broadband (e.g., WiMax) are heading that way as well. Many designers have discovered that the key to more performance (both throughput and range) now lies in advanced antennas.

What do we mean by advanced? Phased arrays are increasingly found in cellular systems. Multiple input, multiple output (MIMO), formerly a laboratory curiosity, is now available in inexpensive production WLAN equipment. A common argument against MIMO has to do with cost (hey, good tires cost more than the discount-store specials); power consumption is also sometimes mentioned. But it's all but guaranteed that 802.11n is going to be based on MIMO, probably with more optional features than any designer will ever be able to comprehend.

Still, if cost and power are issues, there must be something between diversity dipoles and MIMO.

And there is. Atheros just announced its next-generation VLocity chip set, incorporating multiple antennas-digital beam forming on the transmit side and maximal-ratio diversity combining on the other side. The company claims up to 10 dB of gain from this arrangement. Video54 also announced a digital beam-forming solution, said to be very inexpensive; Netgear is using it in a new residential access point. Interdigital has a very inexpensive steerable antenna aimed at both the cellular and Wi-Fi worlds.

MIMO antennas
All of this has led to a nascent debate on exactly what MIMO is. I've found some analysts and press people are now using the term to describe any multiple-antenna implementation. I tend to be a strict constructionist, so I still think of MIMO as an antenna system with multiple unique signals on each antenna, as opposed to just steerability or some other parameter alone. But the definition now seems all but certain to become rather broad. No surprise; this almost always happens in high tech.

And, in the long run, it doesn't really matter. The proof is in the performance, as the radio flies through the turns and accelerates down the back stretch. It's going to become quite clear which designs work, and which belong at the discount store.

Craig J. Mathias (craig@farpointgroup.com), principal at Farpoint Group (Ashland, Mass.).