No need to rush 802.11a

On the surface, it sounds like a dream come true. Everybody wants a faster version of a familiar standard. Unfortunately, current 802.11a products ask you to make tradeoffs you're probably not willing to make.

The new 802.11a standard allows for products that operate at frequencies near 5GHz. Bandwidths can be as much as 54Mbps (or more with some proprietary encoding techniques). Compared with the older 802.11b standard--which is prone to interference from sources including microwave ovens and cordless phones, and operates at a relatively slow 11Mbps--the new standard sounds like a great improvement.

If it weren't for the laws of physics, it would be great. Unfortunately, because the transmission frequency is doubled, the range of the signal is effectively cut in half. To give a large open area comparable wireless coverage, you need four times as many 802.11a access points as 802.11b access points. Considering that the 11a access points and interface cards cost two to three times as much as 11b products, you've got an expensive upgrade.

Worse, the higher frequency of 802.11a radio signals may prevent them from reaching their destination. They may not penetrate walls, floors, ceilings and furniture in the same way that 802.11b signals did, which could leave you without any wireless access at all.

Incidentally, a series of tests have supported the contentions above. The 802.11a access points that we tested, which are available today, had remarkably short and inconsistent coverage.

On the other hand, you get five times as much bandwidth. Not a bad thing, assuming you need it. Even if you do need that much bandwidth for your wireless devices, you may not need it for all of them, or you may need it only for a few specialised devices. For example, most companies find that the slower, but less expensive, 802.11b networking is entirely adequate for access to the Web and email, which is what most employees need. Why? Because most companies have Internet connections via a T1 line, which runs at a slow 1.5Mbps.

The only good reason to deploy 802.11a is if you have a bandwidth-intensive application that must be wireless, such as multimedia development.

But the chances are that if you need it at all, you need 802.11a for only a few people, and for them, you can install the access points they need, while everyone else uses 802.11b. Until the prices come down and the range problems are solved (which may require FCC action), your best bet is to forget about rolling out a large 802.11a wireless network.

Of course, you might feel like waiting for 802.11g to arrive. This standard will bring products that operate on the same frequencies (2.4 gigahertz) as 802.11b, but will support speeds up to 54 Mbps. Unfortunately, the IEEE has yet to agree on the standard, and products that would support it are still in the future. If you need wireless now, this isn't something worth waiting for.

Do you think deploying 802.11a makes sense? Tell us through Talk Back!