Duke Is Investing in a Half-Baked 802.11n Standard

You can never be thin enough, have enough money or a fast-enough wireless connection. That's why everyone wants 801.11n, vendors are still locked in a standards war, and no one should spend any serious money on an 802.11n deployment.

That's why I'm not sure why Duke University plans on deploying more than 2,500 Cisco 802.11n APs (access points). Yes, I know why they want it. I want it too. Who wouldn't want average data throughput performance of nearly 130M bps and about double 802.11g's range?

The fight over the 802.11n has been going on for almost five years now by my count, and it's still not done. As is so often the case in standards, two major vendor sides squared off over which group's approach would become the one true, and, thus money making, standard. At one time, Mitsubishi and Motorola paired together for yet another standard and Qualcomm had its own horse in the race. Fortunately, they bowed out early leaving us with only two competing groups.

On one side was the Task Group 'n' synchronization, or TGn Sync. It counted Intel, Atheros Communications, Nortel, Samsung, Sony, Qualcomm, Philips and Panasonic among its supporters. On the other was WWiSE (World-Wide Spectrum Efficiency). Airgo Networks lead this faction. The two sides reached a compromise agreement. In theory that is. The 802.11 Working Group's Task Group N (TGn) voters still haven't approved the technology as a standard.

Instead, what we have today is the 802.11n Draft 2 'standard.' This is not a final standard. It's also not a standard that works between vendors. So, if I come on to the Duke campus with my Apple MacBook Pro, there's no guarantee that it's going to work at 1T 802.11n speeds with Duke's newly installed Cisco Aironet 1250 Series AP-based WLAN (Wireless LAN). Oh, the MacBook, a Dell, a Lenovo, or whatever laptop you happen to have with an 802.11n card, will work. But it will work by falling back to 802.11g.

I'm not sure that at a list price of $1,299 per 1250 unit, I'm really spending my network money well. After all, if I want 802.11g's theoretical maximum speed of 54M bps, I could buy a decent 802.11g AP, say Cisco's own Linksys WAP54G Wireless-G Access Point for about $70. Of course, if you want to spend 18 times the price of 802.11g for 802.11g service, don't let me stop you. You're the one who will have to explain your decision to the chief financial officer. Good luck with that.

What's that? 802.11n is almost ready to go? A firmware update will make all 802.11n devices work and play well together? Oh dear? Where have you been? 802.11n is still nowhere close to being a real standard. The official date for 802.11n to become a real standard is currently Dec. 31, 2008. It's not going to make that date. Why should it? The vendors fighting over 802.11n standardization have never met a deadline they couldn't blow.

I'm not just being a pessimist. I follow this stuff in my copious free time and Task Force N still hasn't finished up Draft 3 yet. They hope, if they can fix the—OK, now take a deep breath—127 unresolved comments by March, Draft 3 will be passed.  But I'm not holding my breath.

And, let's not forget the real 802.11n killer. Before anything can become an IEEE standard, everyone who has a patent that affects the proposed standard must sign a LoA (Letter of Agreement) that states they won't sue anyone for using their patent in a device deploying the standard. It makes sense. What's the point of a standard where one of the companies that helped make it can sue you for violating their patent? 

Specifically, CSIRO (Commonwealth Scientific and Industrial Research Organization), a research group sponsored by the Australian government, has not signed off on 802.11n. I don't expect them to either. CSIRO has aggressively defended its patents against Wi-Fi vendors, such as Buffalo Technology, and I don't see them co-operating with the IEEE.  Especially when you consider that, when I last checked, Dell, Intel, Microsoft, HP and Netgear are all suing CSIRO to get those patents overturned. Those companies are, of course, also members of the IEEE task force trying to get 802.11n officially blessed and out the door.

What does that mean for you? Well, if you're Duke University, you're not going to change your plans regardless of what I say. Besides, Duke is no run-of-the-mill Cisco customer. The school has been working with Cisco on its 802.11n technology for almost a year now. I strongly doubt that Duke is paying retail for its Cisco equipment.

But, if you're working in enterprise lusting for high-speed wireless, well think again. 802.11n is still a long, long way from being a real standard. From where I sit, there's still too much uncertainty to spend a significant chunk of my IT budget on 802.11n.

Darn it.