Air Time: Wi-Fi and Commoditization

Quick! What do soybeans, pork bellies and natural gas have in common? If your answer involves your favorite local diner, you're way off track. These are major commodities bought and sold on international exchanges every day. Commodities are a good thing; they imply an expectation of quality and availability along with price competition. What's not to like about that?

It's commonplace to hear pundits talk about the commoditization of IT. And while there may be a lot more variation among notebook computers than soybean bushels, the prevailing wisdom is that technology commoditization is also beneficial. However, there are times when commodity technology impedes progress and establishes an ugly legacy with complex challenges. Wi-Fi is the latest example.

All Hail Ethernet!

If commodity Wi-Fi is arguably bad, then an opposing example is Ethernet, which stands as a testament to the power of open standards, the economies of siliconization and the market power of mass adoption. Despite the sage advice once given me by an IBM salesman--that the laws of physics would prevent Ethernet from ever exceeding 10 Mbps--the technology has been continuously enhanced over the years. It dominates not only the LAN but, increasingly, metro networks as well. It's interesting that, in an era of MAC-layer switching, very little of the original contention-based Ethernet specification is relevant. Still, it's the poster child for all the good that comes out of standards and commoditized technology.

While the incredible success of Ethernet speaks well for the value of commoditization, what happens if a technology reaches this stage prematurely, in a form that inhibits future progress? Like any legacy system incapable of meeting current, let alone future, needs, it begs to be replaced by something different, but doing so presents enormous obstacles.Is Wi-Fi the Future of Wireless?

Wi-Fi seems to be taking the same path as Ethernet: starting out slow and expensive, and undergoing continuous enhancements that make it faster and cheaper. Along its path to commoditization, Wi-Fi is penetrating new markets. Originally a niche technology, useful only in certain vertical industries, it has achieved ubiquity on home networks, much to the chagrin of moonlighting cable technicians. Predictably, the mass consumer appeal of Wi-Fi drove down prices for wireless routers and notebooks with embedded Wi-Fi.

The next phase is the mass adoption of Wi-Fi in the enterprise, followed by metro-area Wi-Fi--if you believe those who preach about wireless mesh, bridging the digital divide, and sticking it to entrenched telecom and Big Cable. Herein lies the problem. Unfortunately, Wi-Fi is a poor metro-area wireless technology. It's designed to be a small-cell, contention-based LAN system and, even with myriad enhancements, it's hard to believe you'd create anything like what we have today if you had the luxury of starting with a clean slate.

Charting Wi-Fi's evolution to follow the same course as Ethernet's is probably a mistake; after all, Ethernet has evolved into a high-speed, point-to-point packet-framing and -signaling standard beholden to none of the physical layer media-contention issues of the past. Wi-Fi is different. That IBM salesman may have been wrong about Ethernet, but combine the laws of physics, a LAN MAC and unlicensed radio with large-scale metro Wi-Fi deployments and the result isn't pretty. In fact, it's so ugly that I'm betting a large proportion of current metro Wi-Fi projects will collapse under the weight of the 802.11 legacy.

Yet because Wi-Fi has become a commodity technology, it's driving metro wireless. Arguments that 3G or mobile WiMax are superior technologies fall on deaf ears. All that's left to do is find a way to make Wi-Fi work. Unfortunately, I'm not optimistic. Dave Molta is a Network Computing senior technology editor. He is also assistant dean for technology at the School of Information Studies and director of the Center for Emerging Network Technologies at Syracuse University. Write to him at [email protected]