* Kicking off our Wireless LANs Technology Tour
This week, we kicked off our Network World Wireless LANs Technology Tour. Keynoter Craig Mathias, principal at Farpoint Group, brings to the stage his wide-ranging expertise in all things wireless. I discussed the state of the industry with Mathias and some of that interview was carried in Network World (link below).
Below is more of that interview, including a look at wireless standards and what considerations IT managers should take in rolling out wireless networks. To register for this tour, visit: https://www.nwfusion.com/events/wlan/index.html
Q: The technology behind wireless LANs is evolving rapidly. What do you think has been the biggest change over the past year?
The replacement of 802.11b with 802.11g. 802.11g is, of course, backwards compatible with .11b, and now .11b-only products are becoming hard to find (although quite cheap as they’re closed out). More performance is always a good thing. This is not to say, by the way, that 802.11a isn’t going to see increasing use as well. .11a and .11g use essentially the same technology and have similar performance, but the far greater bandwidth available to .11a means that its effective performance will be much greater than .11g in most cases. Of course, we’re now seeing combined .11g/.11a adapters that will eliminate the requirement that a user knows which technology is being used. Keep in mind also that .11n is under development, with throughput likely to be at least 108M bit/sec. There really doesn’t seem to be any upper boundary on WLAN throughput at this time.
Q: Where do you see the biggest push for WLAN technology? Switching? Access points? End devices?
All three, really. Client devices always need to be smaller, cheaper, faster, and with improved power consumption characteristics. Access points are going to become “thin” and very inexpensive, so device integration is a key requirement here. And switches will have even more functionality, essentially becoming very-high-performance specialized computers not unlike routers, with similar sophisticated management capabilities. As is often the case in any high-tech product, software for management and value-added features is going to be the differentiator.
Q: Is wireless networking a one-size-fits-all proposition or are there considerations IT organizations need to take in deciding to roll out WLANs?
It’s most definitely not one-size-fits-all. While the technologies involved are identical, network configurations vary widely. Most WLANs today are based on “fat” access-point architectures, wherein each AP is separately managed, and all control functions are distributed. Newer switched architectures centralize management and control, and many lower-cost (and even low-cost) switched products have recently been introduced, meaning the range of choices is much greater today. Planning a wireless network is not that different from building a wired infrastructure. Once key goals are determined, the right combination of equipment can be selected quite easily. It’s also important to note that, just like wire, wireless configurations will be modified, upgraded, and otherwise changed over their life cycle. Network management capabilities, which differ significantly between products, are probably the place to start now – especially in determining how your operations staff and systems will mesh with the proposed wireless network.
Q: The standards have been coming fast and furious in wireless networking. IT organizations are tempted to wait for things to settle out. What advice would you give them? What should they standardize on?
There’s really no such thing as “settling out” when it comes to standards or their underlying technologies. For example, just look at 802.3. There’s been a steady flow of improvements throughout the history of Ethernet, resulting in the need for upgrades and even wholesale replacement of equipment from time to time. That’s just the way technology works, and we need to expect this in wireless given the relative immaturity of the industry and the ever-increasing number of players in this space. So we could behave like the guy who argues that PCs are always getting faster, cheaper, and better, and therefore never buys one, or look at it, again, from an ROI perspective. If we can get a good return on the investment given the applications and technologies of today, and we can manage the cash-flow issues (getting easier all the time, because WLAN prices continue to fall while price/performance continues to improve), then now is a great time to address appropriate applications with WLANs.
With respect to specific standards, we encourage enterprise users to adopt 802.11a, because of the huge amount of bandwidth available to it. More bandwidth is like a highway with more lanes – it’s easier to go, and go fast, when you don’t have a ton of slow traffic in front of you. For some reason, the backwards-compatibility argument of 802.11g appeals to us, probably because backwards compatibility has always been essential in the wired world. It’s not, however, in the wireless world. And over-the-air interoperability isn’t important, since we have interoperability over the backbone network. With respect to security, many are waiting for 802.11i, with its much stronger AES airlink encryption. But this isn’t essential if one uses a VPN or some other form of upper-layer security. So I think we have everything we need in terms of standards-based functionality today, although we’ll be adding higher throughput in .11n and other capabilities like mesh networking over time. There’s no reason to believe that 802.11 won’t be extended with new capabilities for the foreseeable future – just like Ethernet.
Q: What do you think will be the biggest change in the industry in the next year? How should IT managers prepare for this change?
Higher performance – with proprietary or pre-standard 100+ M bit/sec solutions appearing in advance of the ratification of 802.11n. More performance is always desirable, even for traffic that doesn’t require it, because more performance also means more capacity – even low-speed traffics moves more efficiently on a high-bandwidth network; we’ll also need more bandwidth to provide greater headroom for time-bounded traffic like voice over Wi-Fi. But users will need to be on the lookout for possible (and perhaps likely) incompatibilities between pre-standard products and what eventually is in the standard. As always, the Wi-Fi Alliance will be providing interoperability testing, so we’ll have some (but not absolute) assurance that everything will work together once the standard is approved and compatibility tests are in place.
