Changing the Face of Wireless

New Rules

In the mid-1990s, Greg Raleigh and V.K. Jones were exploring the vagaries of wireless communications at Stanford University when they discovered a way to turn multipath on its head. As if in some parallel world where up is down and strikes are balls, they discovered that multipath isn't bad. In fact, if your goal is to transmit lots of data over a wireless network at maximum distance, multipath is good. In a discipline defined by Shannon's Law, the mathematical rule governing channel capacity, the rules suddenly changed.

The discovery came to be known as MIMO, short for multiple-input, multiple-output. Rather than using a single radio, MIMO uses multiple transmitters and receivers on the same channel, thereby increasing channel capacity proportionally. Spatial multiplexing was born, aided by advances in digital signal processing. Raleigh and Jones commercialized MIMO at Airgo Networks and began selling their chipsets to home wireless router manufacturers, which had grown wary of product returns from customers who couldn't get a good signal in the upstairs bedroom. Airgo's technology solved those problems and boosted performance.

Competitors, caught off guard by Airgo, created market confusion, mostly by applying the MIMO label to a range of technologies. Nobody owned the acronym, so technically speaking, they weren't lying when they called their systems MIMO, but the benefits of these competitive designs were modest.

The Latest BattlegroundAs spatial multiplexing became accepted as "true" MIMO, the battle shifted to the hows and whens of the IEEE standards process. The 802.11n task group was charged with developing a new physical-layer standard delivering throughput of at least 100 Mbps.

Two competing proposals emerged. Airgo teamed with Broadcom, Conexant, Texas Instruments and others to form WWiSE. A group known as TGn Sync, led by Atheros, Intel, Agere and Marvell, offered an alternate plan. The battle of wireless silicon developers had begun.

Spirited competition is often good, but in the IEEE, it makes it almost impossible for any proposal to garner the 75 percent support required to establish a standard. After much wrangling, the EWC (Enhanced Wireless Consortium)--comprising all the major 11n silicon players except Airgo--was announced Oct. 10. EWC also lined up support from major enterprise and consumer wireless equipment providers, including Cisco, D-Link, Netgear and Symbol.

Assuming EWC can push its proposal through the IEEE, this may not be such a bad thing for Airgo. Yes, the company was making money selling lots of prestandard MIMO chips and would have been the prime beneficiary if the 11n standard had been delayed. However, after three generations of MIMO chip development, Airgo still plays a technical leadership role. That's not to say its competitors won't catch up, but if 11n takes off as many expect it to, that rising tide will lift plenty of boats, including Airgo's.

Dave Molta is Network Computing's senior technology editor. Write to him at [email protected]1022