Meeting Tomorrow's WLAN Needs

No one can predict how much more bandwidth we'll need on wireless networks. But it's all that hard to assume "plenty."

Dave Molta

April 9, 2004

2 Min Read
Network Computing logo

Build for the Future

Preserving the sanity of many wireless network managers is the fact that most of today's WLANs are substantially underutilized. With the possible exception of health care, the industries that have embraced WLANs, including education, government, retail and logistics, make modest demands on their wireless bandwidth. For emerging horizontal enterprise and hotspot use, user density is relatively low and the most common applications--Web access and e-mail--are bandwidth-efficient.

Although performance isn't a big issue today, it will be tomorrow, leading to the same provisioning dilemma as with Ethernet. With wireless, however, technical obstacles are substantially more complex. Before we can get fatter wireless pipes, we must consider performance, capacity, cost, range, backward compatibility and other factors.

IT pros place a high value on backward compatibility, and Ethernet designers often get it for free. Not so with wireless. A NETWORK COMPUTING survey suggests the enterprise wireless growth path of choice over the next three years will be 802.11g, which theoretically delivers about five times the performance of 11b with full backward compatibility. However, effective throughput in a mixed 11b/11g environment will be substantially lower. And though b/g delivers the best range today--an important consideration for some environments--long-range radio transmission is exactly what you don't want.

Think MicrocellsThe best architecture for a scalable WLAN incorporates small coverage cells, which diminish contention. With 2.4-GHz b/g, that's hard to do. There are only three nonoverlapping radio channels in this unlicensed band, which limits channel reuse and overall capacity. Interference from other wireless devices is also common. Even if you optimize your access points to support small cells, it's almost impossible to control the transmit power level of client adapters. That leads to co-channel radio interference and diminished throughput.

A better design is a microcellular 5-GHz 802.11a network with a b/g overlay. Think of the b/g portion as your general-purpose wireless network. Although most users will find that system adequate, the 5-GHz 11a system gives you more scalable capacity, a premium WLAN environment that will deliver enhanced performance today and insurance for tomorrow. Still further into the future, you may find it necessary to shrink cells even more to support next-generation 100-Mbps WLAN standards like 802.11n. Increased allocation of unlicensed spectrum in the 5-GHz band will likely facilitate this future third service layer.

The incremental cost of this new design, especially in today's highly competitive WLAN switch market, is modest. But most important, it provides you with a more scalable environment, one that supports both legacy and emerging standards.

Dave Molta is a NETWORK COMPUTING senior technology editor. Write to him at [email protected]

SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox
More Insights