Network Computingwww.networkcomputing.com

Network standards are great, but they can have a negative effect on innovation. Standards stimulate product commoditization and price competition, which leads vendors to skimp on R&D. The problem is especially acute when it relates to technologies in early stages of evolution, such as wireless LANs. If standards get locked in too early, they can become an albatross.

Fortunately, there still appears to be enough gold in sight to encourage investment and innovation in wireless at nearly every step of the value chain. Airgo Networks, a start-up in Palo Alto, Calif., is looking to innovate within the 802.11 standard. The company recently began sampling its AGN100 Wi-Fi chipset. The system relies on a smart-antenna signal-processing design called MIMO, originated at Stanford University by the company's principals. The technology promises dramatic improvements in range, throughput and reliability. In a market of me-too silicon, it's nice to see something different.

It's not yet clear whether the performance enhancements are much different from other proprietary channel-aggregation features that theoretically double 11a/11g performance to 108 Mbps. (You're actually lucky to get a 20 percent increase in throughput.) But the new product's range benefits may draw significant interest. While I've long promoted the benefits of short-range wireless systems, which diminish contention and facilitate channel reuse, there's no question that the range limitations of today's WLAN offerings constrain the market for many emerging and appealing high-bandwidth wireless applications, like video.

Although Airgo's performance enhancements are proprietary (the company reportedly plans to submit them for consideration as a standard), its new antenna design is an example of innovation within the 802.11 standard. The company is not altering the PHY or MAC designs; instead, it's just taking advantage of improvements in signal processing and RF antenna technologies and applying them to wireless networks.

One key question about the technology relates to how much the benefit is reduced when one end of the connection is Airgo and the other end uses Atheros, Broadcom or Atmel chipsets. It's highly likely that the system is optimized for Airgo on both ends. For some applications, such as wireless home-entertainment systems, this quasiproprietary approach is acceptable, so long as costs stay low. But if this technology proves valuable for wireless business applications, you know what everyone will be asking: Can we find a way to standardize this technology?

Post a comment or question on this story.