Chipmaker Intersil and networking giant Cisco announced last week that they are releasing a reference design, sort of a basic recipe, for building wireless network interface cards based on the draft IEEE 802.11g standard.
The idea is that wireless LAN equipment vendors will take the Intersil-Cisco design, which exploits Cisco software features for wireless clients and for advanced security, to build interface cards. The cards, in various sizes, would fit into laptops, PDAs and other devices and use a radio link to connect with a wireless access point.
The deal is the latest effort in the 802.11g area, which is touted as a way to help enterprises shift from current relatively slow 802.11b wireless LANs to faster nets that still operate in the same 2.4GHz band. Advocates of 802.11g say it offers enterprises an easy way to migrate their existing 802.11b radio infrastructure, including interface cards and access points, to a higher speed.
Intersil is one of the chief backers of 802.11g, along with Texas Instruments. Cisco seems to be covering all wireless bets, with plans to use 802.11g chips, when available, in a line of products aimed at existing, or expanding, 802.11b networks.
But the main question is how, or whether, 802.11g truly fits into enterprise plans. For one thing, a lot of enterprises don't yet need a migration. Current installations of 802.11b, which has a theoretical data rate of 11Mbps, are typically few and small, and a lot of enterprises haven't yet decided to deploy wireless, or to deploy it widely.
Second, wireless LAN products based on 802.11a, which uses the less crowded 5GHz band and has a theoretical maximum data rate of 54Mbps, are just now starting to appear. And many more, including 802.11a products based on Cisco's own Radiata chips, are due in the middle of the year.
About the same time, the Wireless Ethernet Compatibility Alliance (WECA) should have an initial test suite available for certifying interoperability among different brands of 802.11a products. By contrast, the very first 802.11g products aren't expected until later this year, with no announced plans for interoperability testing.
Third, there is the question of what distances the various standards support. Higher frequencies have shorter ranges. In theory, to cover the same area as an 802.11b net today, you'll need four times the number of 802.11a access points. Networks based on 802.11g should, again in theory, give you the range of 802.11b but something closer to the throughput of 802.11a.
In reality, however, it all depends. You have to take into account your actual site layout, the materials in the buildings, how furniture is arranged or inventory is stacked, actual location of users, what kind of traffic they generate, and more. In other words, the theoretical limitations of 802.11a, or 802.11b for that matter, may or may not prove to be real limitations in your particular deployment. Advances in antenna design will also boost range by creating still more sensitive, efficient, and effective transmissions and receptions.
Cisco's approach seems to be that whatever customers want or need, they can have it. If you want 802.11b, buy it from us. If you want fast 802.11b, buy our future 11g products. If you want 802.11a, we've got that covered too.
The problem with all this -- at least for the vendors -- is that it doesn't amount to an enterprise wireless strategy. In fact, it's an implicit admission that, basically, no-one knows what the heck network executives are going to decide about wireless LANs.
And that's because network executives are still figuring out what they can, and can't, do with them.
The 802.11b products today are cheap to buy, pretty easy to deploy and usually very interoperable. They offer adequate bandwidth for a wide range of applications. The security problems are real, but known, and can be managed.
The 802.11a products promise considerably more bandwidth, at an affordable price increase. And they should scale to support much larger groups of users.
It is possible to figure out the business case for today's products, evaluate them, and test, measure and secure them. The new type of wireless LANs, when they become real, will go through the same process of being evaluated in terms of real business needs.
So 802.11g may or may not have a future. The same is true of Bluetooth and ultra-wideband wireless.
Network executives don't care. And they don't have to. They only have to care about whether wireless LAN products, real ones, can pay off for their business.