Good idea? Would you take a risk on draft 802.11n? Have any questions for the folks at Morrisville on how they're doing this? Here's the place to post.
Re: N.Y. state college takes the plunge, plans 802.11n WLAN rollout this summer
The best place to talk about enterprise networking.
|
|
What about 11b and 11g?
Unless these are tri-radio APs, how can they deliver b/g service on a 20 MHz-wide 2.4 GHz channel, 802.11n service on a separate 40 MHz-wide 2.4 GHz channel, and 802.11n service on a 5 GHz channel?
Or are you assuming that they'll deploy just one 40 MHz-wide 2.4 GHz channel that serves 802.11n clients and can serve legacy clients in compatibility mode? That will, of course, reduce the experience of 802.11n clients.
Re: What about 11b and 11g?
The Meru AP300 line of APs supports 2 radios. Both radios can operate in either the 2.4 or the 5GHz band and can operate in the in 802.11n mode or 802.11 a/b/g/n mode. Morrisville is planning to deploy 1 radio on the 5GHz channel to serve purely 11n clients. The other radio will be on the 2.4GHz channel to support both legacy b/g clients on 20MHz as well as 11n clients
Kamal, I think the issue is
Kamal, I think the issue is this: if an 11n user associates with the AP300 in the 2.4GHz band, what happens when a legacy 11b/g client associates on the same band? How does the AP handle this, and what's the effect on performance for both the 11bg and 11n clients?
Thanks.
And...can these different
And...can these different clients (11bg on the 20Mhz channel, 11n on the 40MHz channel) access the AP at the same time, on these separate channel? Or is it ONLY the 40MHz channel that's "active"?
b/g will override
My understanding is that if b/g clients associated to a "n" access point everyone will drop to the lowest common standard, i.e. 802.11b. Only if all client are using 802.11n will you get the full speed. There is no point using 2.4GHz 802.11n if you have more than on AP (or channel blanket with Meru) co-located as you can't co-locate a second cell as there is only one channel spare. Only use 802.11n in 5GHz would be my advise, and only support 802.11n clients if you are going to do so. Of course anyone deploying pre-standard on such a scale wants their head checking in any case.
One last point - 900 APs for 1800 students? Who has surveyed that? I think someone has stuck an extra "0" on there, 90 maybe, never 900!!!!
price
What impact will 11n's premium price tag have on near-term deployment decisions?
As the story mentions, the two-radio Meru 11n access point has a list price of $1,495, the controller (with slots for 5 boards, each handling up to 200 access points), is about $65,000. The access point price is about twice what Meru charges for its two-radio 11abg device. On the other hand, the net capacity is certainly boosted by more than twice that of a conventional wireless LAN.
Price
I certainly don’t want to try to predict what others may do but I am happy to share our thoughts with you. First of all, we always expect to pay a premium for technology products that are the newest, fastest, and greatest as that has always been a fact of technology life. In addition to looking at the cost of a technology product, we feel it is critical to consider the remaining useful life of a product and even more critical in the wireless world since the rate of change is staggering. Given the current and expected throughput needs of our 3,200 students and given that we believe the 802.11a/b/g products are very near end-of-life, our decision to implement 802.11n (draft) was the most fiscally sound decision for us.
Is this really a good idea at this time?
I just wonder about all that infrastructure upgrade, regarding costs of materials, labor, training, and support...
The article mentions a student body of some 1,800. with 900 access points deployed, that is averaging two students per access point, plus all the other costs beyond just the APs.
There's no way I would approve such a project if it were left to me since we're not even dealing with a ratified standard. This seems somewhat irresponsible and if I was paying tuition here I'd wonder where else my money was going.
Is this really a good idea at this time?
I can understand your concerns. However, we have been a wireless campus and our students have been using wireless since 1999, so the decision for us wasn’t whether to go wireless or not, it was what to replace our 802.11 FHSS wireless product with. You are also right, there is a risk in deploying a product before the final standard is ratified. For us, this risk was definitely mitigated by the Wi-Fi Alliance’s decision to certify 802.11n draft 2.0 products and by the large number of companies producing 802.11n draft 2.0 compliant products, expecting only firmware upgrades to the final ratified standard. Is there still a risk? Sure, but our decision to deploy 802.11n (draft) is the most fiscally sound decision we could have made at this time.
Questions
I was wondering what your projections for students connected to a single AP is for the dorms are where there is likely to be the highest concentration of students would be?
Additionally, most of the dorms are older buildings that are not wireless friendly regardless of running at 2.4Ghz or 5Ghz, how do your testing plan to simulate typical student experience in the dorms?
How is the rolling deployment and upgrade of APs affect the bottom dollar cost? From what I read you are basically going to pay for an A/B/G network only to turn around and replace it.
Being that you are planning on using PoE, is there a basic infrastructure for this already? As it was mentioned the college recently already upgraded hardware to Gigabyte so is that now going to have to be replaced? And the current Raylink APs have a dedicated wall socket already with the increase in range and performance of 802.11n I would think you could pick and use about half the locations currently occupied by the Raylink APs.
Being a former Morrisville student and a current Alumni how is the cost of this new network translating to student expense increases?
I am all for seeing Morrisville finally rid itself of 802.11 I believe it was well past it's prime during my attendance between 2001 and 2005. My concern is that the foresight and planning of the current wireless network seemed extremely lacking from my experience using it for 4 years and I don't want to see the same happen with it's replacement.