Multiple Input Multiple Output technology is capable of turning the worst enemy of wireless network planners — multipath interference — into one of their strongest allies. While doing research at Stanford University in 1996, Greg Raleigh and VK Jones discovered how the multipath reflections created by radio transmissions hitting objects in their paths could be turned into separate, virtual transmission channels guided by smart antenna architectures. Though the MIMO concept is now 9 years old, it's just starting to be viewed as central to future Wi-Fi and mobile standards. Telephony's Dan O'Shea recently spoke with Raleigh, who along with Jones, runs Airgo Networks, a developer of MIMO chipsets.

On MIMO's early days: In 1896, during the first demonstration of wireless transmission, Guglielmo Marconi had to send a signal over a hill, and even then, there was this problem with multipath, with the signals reflecting off things and creating interference as they passed through the environment. For the next 100 years, people tried to undo what nature was doing to the wireless transmission. In 1996, I wrote a paper stating that multipath can have multiplicative capacity properties. Some people refuted what I said, and other people wrote their own papers, but by 1999 or so, mine was commonly accepted. Now, MIMO is the top research topic among wireless Ph.D.s. Everybody in wireless understands this now. They didn't understand it nine years ago.

On former Raleigh/Jones venture Clarity Wireless: We wanted to do MIMO when we were at Clarity, but we couldn't figure out then how to get it on a chipset. We did smart antennas at Clarity, which is different because it tries to increase capacity by using a single signal. MIMO uses multiple signals sent from multiple antennas. After two years, Cisco Systems bought Clarity for its fixed wireless effort. We started Airgo and spent nine months cracking the code for how to get MIMO on a chipset. We had our first chipset in 2004, we're now launching our second generation chipset, and we will have the third generation sometime later this year.

On the 802.11n Wi-Fi standard in the works: People told us not to go after Wi-Fi first because the market was too crowded. We brought MIMO first to Wi-Fi because of the nature of the market. There is a lot of rapid growth and innovation, and companies making decisions very rapidly. MIMO is the basis for the proposals at the heart of the 802.11n standard, for which there should be a draft by the end of this summer. There were two competing proposals, but they both use the same core technology, which is MIMO. 802.11n will be going commercial sometime next year. Then, you will see MIMO in all kinds of consumer electronics products, including TVs and set-top boxes.

On the potential for MIMO in the cellular industry: Essentially, MIMO sends multiple signals on the same frequency using the same spectrum, so you can get five times the bandwidth on the same spectrum, which is important for the cellular industry during a time of fluctuation in the value of spectrum. There are going to be 270 million devices out there by 2007 sharing the spectrum. In cellular, spectrum efficiency is everything. MIMO also will be the only way that cellular will be able to deliver DSL-like speeds. In the cellular industry, 3GPP is looking at this, and we're joining 3GPP. We went to Wi-Fi first because it's the market that moves quickest, but cellular will come later. There could be a cellular standard in 2007.

On WiMAX: We see no reason why service providers should spend money on that equipment. WiMAX is an unproven market, and we don't see a business model. There's no real business in competing with DSL and cable and cellular. I've lived through this already when Cisco bought Clarity and shut it down because the fixed wireless market didn't pan out. WiMAX will be a market for MIMO if it turns out to be a market.

For insight on service quality management, watch our Webcast featuring OSS Observer's Patrick Kelly.
WWW.TELEPHONYONLINE.COM