FWIW -- I think we are stuck in a vicious cycle that prevents the growth of 802.11a. LAN admins don't provide 802.11a APs because there are too few clients to justify it. Laptop buyers don't by 11a because there's nowhere to use them. My perception is that most "high-end" laptops have come with a/g cards for a couple years now, but few of our departments and even fewer students are getting high-end laptops. I believe that an a/g card as standard equipment is less common now than it was 12 to 18 months ago. The combo card was standard equipment on "mid-grade" laptops for awhile, then they disappeared. It may be within our power to break that cycle by building the infrastructure and education our users, but don't expect it to happen quickly.
I agree with all the potential virtues of 11a mentioned already, but I don't think there are enough 11a users on our network to justify much additional cost. The cost difference between a g-only and an a/g WLAN is trivial either. Considering the addition cost of dual-band APs, as well as pig-tails & antennae (diversity on both radios, of course), an a/g WLAN cost about 45% more than g-only. Whatever the cost difference, you have to decide if it is best spent providing 11a coverage where 11g is already available, or if it should be spent on something else - providing more 11g coverage, for example. Without a crystal ball to consult, I think we also have to consider that we may be installing 11n devices before 11a VoWLAN sets and other similar devices are commonplace. >From an engineering standpoint, I think 11a is great. In a perfect world the question would be, "should I bother to provide 802.11g?" I also agree with Dave's assessment that if you can get 20%+ of your users on 11a it's worth it. Short of that, however, you should consider if there are better things to spend your money on. Chuck Enfield Sr. Communications Engineer PSU, Information Technology Services Suite 110, University Support Bldg. 2 University Park, PA 16802 ph. (814) 863-8715 fx. (814) 865-3988 -----Original Message----- From: Phil Raymond [mailto:[EMAIL PROTECTED] Sent: Thursday, February 23, 2006 5:10 PM To: [email protected] Subject: Re: [WIRELESS-LAN] 802.11a People will look back at 802.11b similar as they look back at using a hub today - yuck. Why buy a hub when a switch offers better performance at a decent price point? Same story for 802.11a. It offers better performance simply from the fact that there is a lot more bandwidth/channels. By the end of 2006 any high end laptop will be called a/g, and VoWLAN handsets will start with a/g offerings. The chipset cost difference is minimal between b/g and a/g. When I would go to trade shows the last couple of years I would always require wireless users at our booth to use 'a' instead of 'b'. With all the 'b' AP's nobody could get wireless access, yet we would be screaming on 'a'. I'm gonna miss those days as more applications are enabled on 'a' and users are migrated there....whether they like or realize it. What people are starting to realize and has been an industry secret by the AP manufacturers, is that 3 channels does not make for a robust WLAN. The shared bandwidth that occurs really brings down the aggregate bandwidth. -----Original Message----- From: Chris Hessing [mailto:[EMAIL PROTECTED] Sent: Thursday, February 23, 2006 3:49 PM To: [email protected] Subject: Re: [WIRELESS-LAN] 802.11a I agree strongly with Dave Molta on this one. At the University of Utah we have been deploying 802.11a right along side 802.11b/g. Currently, our 802.11a usage is less than 1%. I believe that this is mostly due to the fact that you can purchase a b/g wireless card from computer stores for as little as $3 on sale. Along with a lack of education on the advantages of the 5 GHz band. However, when you purchase a new notebook that is wireless capable (which just about all notebooks are these days), it is a small cost to upgrade to a card that supports 802.11a. As for additional support headaches. They should be minimal. 802.11a and 802.11b/g all operate pretty much the same way from a user perspective. The only support problem that I can see is with poorly written drivers (which is a problem with any networking device), and the extremely small chance that a user comes in with an 802.11a only card and complains about the range/coverage. But, I have not seen an 802.11a only card for sale in a long time. On Thu, 2006-02-23 at 16:20 -0500, Dave Molta wrote: > I personally have pretty strong feelings about this issue and feel some > frustation that too many organizations adopt a perspective of choosing > between 11a and 11g. My view is that supporting both 11a and 11g provides > you with more wireless capacity and better performance at only a modest > increase in cost, both on the AP and on the client. > > While there are certainly benefits of engineering your systems for full 11a > coverage by deploying AP's in a dense configuration, even if you choose not > to do that, you get benefits, as long as a reasonable percentage of users > have 11a on their clients. At Syracuse, the University-standard notebook > computer that is made available to students comes with ag support. I think > the incremental cost of 11a from Dell was on the order of $10. Although I no > longer work in central IT, if I did, I would be working closely with the > Purchasing Department to insure that all institutionally-purchased notebooks > included 11a support. I don't think it will be too longer before all > Centrino notebooks come with ag support by default. > > With respect to support, this is largely transparent. In most cases, clients > will attempt to associate first to 11a and roam to 11g if necessary. There > are definitely some latency issues associated with roaming and some client > adapters handle this better than others, but fast roaming is not usually a > huge issue for notebook users, who usually need portability/nobadicity > rather than true mobility, as might be required with wireless VoIP handsets. > > > I'd love to hear arguments from people about why supporting 11a is a bad > thing. It just looks like such a win to me, I don't know why everyone > doesn't do it. Even if you only offload 20% of your client traffic to 11a, > all of those users get better performance and you've also made things better > for the 11b/11g users by offloading that traffic. > > dm > > > -----Original Message----- > > From: Daniel R Jones [mailto:[EMAIL PROTECTED] > > Sent: Thursday, February 23, 2006 3:19 PM > > To: [email protected] > > Subject: [WIRELESS-LAN] 802.11a > > > > CU-Boulder is significantly expanding wireless in student and > > academic areas. The question has been raised about support > > of 802.11a. Even though our new access points support > > 802.11a it may not necessary make sense to deploy the technology. > > > > For those who have adopted 802.11a could you answer the following > > questions: > > > > 1) How much usage of 802.11a do you have vs 802.11b/g? > > > > 2) Do you have coverage of 802.11a in all locations where you > > also have 802.11a or is it provided for specific applications? > > > > 3) Has 802.11a generated additional support calls? > > > > Regards, > > > > Dan Jones > > University of Colorado at Boulder > > > > ********** > > Participation and subscription information for this EDUCAUSE > > Constituent Group discussion list can be found at > > http://www.educause.edu/groups/. > > > > ********** > Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.
