The 802.11 wireless world has evolved to the degree where significant numbers of users are more than happy to bid adieu to wired Ethernet and patch cords, because wireless is now that good. Even better things are due with the arrival of 802.11ac.
Yet even as wireless networks get better in every way, from radio performance to security to on-boarding clients in complicated environments, a smattering of devices lurk in the wings just waiting poop the party. I'm sure I'm not alone in my lament; anyone who runs a large network will likely identify with at least some of the following examples. And, no, this is not an Apple-bashing session. The computer giant has acknowledged its own limits with Bonjour and AirPlay, and other vendors are trying to make up for Apple's slack. Beyond that soap opera, there are other forces at work that disrupt a progressive business WLAN.
Before I start, here's a little background. I'm at the wheel of a large university wireless network with 3,500 access points. The WLAN supports approximately 15,000 students as well as vendors, faculty and staff.
One pervasive problem is that faculty and students want to walk into a classroom and display content from their client devices via the very expensive projectors that most classrooms have installed. Unfortunately, business projectors are guilty of multiple faults: They tend to require proprietary software on each client device (no apps yet, either), they use the forbidden ad-hoc style of wireless networking, they are restricted to 802.11g or lower data rates, and they lack enterprise wireless security capabilities. All of the same issues usually apply to "wireless video gateway" devices that promise to connect you to a projector.
Then there are the utility devices (temperature probes, thermostats, in-car video transfer systems, robots, and so on) that are sold to facilities folks, police, and researchers. The pitch goes something like, "We can sell you this cool thing, and it'll work on wireless networks!" While true, the vendor also has to provide a wireless access point as part of the "cool thing," which just can't be shoe-horned into a dense WLAN no matter how hard you try. "But all we need is a spare channel!" they say. Yeah, there are lots of those to be had in 2.4 GHz, where only three channels can be used. And forget about bringing up a new SSID for every one-off device that comes to campus.
While facilities folks and researchers are willing to stand by my requirement that utility devices must work on campus WLAN, other issues typically pop up. Like devices that only support pre-share-based security, as you'd use at home. Or devices that need static IP addresses sit on networks where we only use DHCP.
Typically, we can get creative to accommodate these users, but many of the devices I'm talking about haven't had their wireless drivers updated for several years. If you're going to be in the wireless client business, selling business-oriented devices to clients that have business-class wireless networks, then your client hardware ought to be compatible with those networks by now. There is no excuse for selling client devices with wireless drivers that are five to 10 years out of date.
The worst of the worst of these incompatibility problems comes when a wireless client requires the oldest, slowest data rates allowed by the original 802.11 standards. WLAN best-practices call for the lowest data rates (1 Mbps and 2 Mbps) to be disabled to improve wireless cell quality and performance. But recently, I've had to do configuration gymnastics to accommodate the biggest ticketing company in the United States (which handles tickets at several campus venues), and a scientific instrument maker whose equipment is beautiful--except for its wireless implementations.
The ticketing company is the poster child for industry players that need to pull their technical heads out of 1992 and get with the program. How many venues have invested in excellent wireless, only to have dog-slow, dated ticket scanners at each gate bringing the overall speed of the venue down? At least in the case of the of the scientific equipment company, I was able to plead my case, which resulted in new driver code a week later that eliminated the legacy data rate requirement, and even brought support for enterprise security protocols. If only every wireless device maker or vendor would follow that example.Lee is a Wireless Network Architect for a large private university. He has also tought classes on networking, wireless network administration, and wireless security. Lee's technical background includes 10 years in the US Air Force as an Electronic Warfare systems technician ... View Full Bio