802.11n on 2.4 GHz

[mchockalingam]

I am seeing a lot of smart phones (including iphones) and laptops connecting using 802.11n radio on the 2.4 GHz band. Is it a good idea to only offer support for 802.11n on the 5 GHz band and turn off support on the 2.4 GHz band?

[weterry]

I don't see a problem with having 802.11N support for 2.4Ghz....  the bigger problem is when someone tries to use 40-Mhz channels with 2.4Ghz...

I saw someone else do this same thing though (disable .11N on 2.4) without explanation. What is your concern or reason for wanting to do so?

-Wesley Terry

[dmantill]

I wouldnt disable 2,4

The main reason is because many of those the smarthphones works ONLY on 2.4Ghz

Therefore even if you have 5. Ghz enabled and you have 802.11n those clients wont be able to associate.

Thats my perspective

[mchockalingam]

I am not disabling 2.4. I only want to disable n on 2.4 and only allow b and g. I want to allow a and n on 5. I have not configured channel bonding on 2.4 since there are only 3 non-overlapping channels however I have channel bonding configured on 5.

With that said, is it better to allow only "b" and "g" on 2.4 and "a" and "n" on 5?

Meena

[weterry]

I'm not aware of any kind of best practice definition that calls for disabling 802.11N on 2.4Ghz.   That is why

I was asking what you might expect to gain from it. If you could explain what you were thinking was a good reason for it, maybe we can help validate that thought.

-Wesley Terry

[mchockalingam]

I was watching a recorded webcast "Delivering Enterprise-class wireless performance with 802.11n" and on that they published baseline test results and this is what it says

Baseline .11g and .11a performance (Mbps) with Band Select enabled

Intel 2915ABG client on 11g had a throughput of 23.745 Mbps

Intel 2915ABG client on 11a had a throughput of 23.762 Mbps

Baseline .11n performance (Mbps)

Intel 4965AGN client with 11n, 20MHz, 2.4GHz had a throughput of 92.99Mbps

Intel 5300AGN client with 11n, 40MHz, 5GHz had a throughput of 169.240 Mbps

Simultaneous .11n performance (Mbps)

4965AGN, .11n, 20MHz, 2.4GHz - 90.77Mbps

5300AGN - .11n, 40MHz, 5GHz - 135.017 Mbps

Simultaneous, same-channel .11g and .11n (Mbps)

2915ABG, .11g, 20MHz, 2.4GHz - 8.3 Mbps

4965AGN, .11n, 20MHz., 2.4GHz - 29.33 Mbps

This is where I got confused. The webcast says

Both radios can operate on the same channel simultaneously but gets an enormous hit on the throughput. The recommendation is "if you are deploying 802.11n, leave 802.11g alone" and "if you are using the same access point channel map as before, move the .11n traffic to 40MHz channel on 5GHz."

I am assuming that this means, do not enable support for 802.11n on 2.4. Is my assumption correct? I am also confused with the second set of baseline data above where simultaneous .11n performance on both 2.4 and 5 are better if they use different channels. How do I make sure that g and n do not coexist on the same channel?

Meena