Wireless: 5GHz vs. 2.4 GHz
 

I have an Airport Extreme in bridged mode. The routing is done by a Buffalo router running DD-WRT.

I set up the AE so that the 5GHz radio had a separate SSID from the 2.4GHz radio. The 5GHz was 802.11n only and the 2.4GHz network was b/g/n.

When my iPad was connected to the 5GHz network, throughput was incredibly slow. When I connected to the 2.4GHz network it was noticeably faster.

So, I've gone back and put the AE's radios in Automatic mode and things seem much better for all wireless clients.

Any ideas why this would be?

Only a couple of my neighbors have wireless and all of them are on 2.4GHz.

My iPad consistently showed better signal strength on the 5GHz network than on the 2.4GHz network.

I'm baffled.

Which model iPad?

Trevor

The "New" iPad (Retina Display).

I should emphasize, though, that 5GHz performance was also poor for an Apple TV (2) and an Airport Express (the recently discontinued one).

The ATV2 is running XBMC. The Express has a Raspberry Pi hanging off it which is also running XBMC. I got excessive video buffering when streaming to both XBMC clients over the 5GHz network.

How far are you from the base station? 5GHz does not penetrate walls, floors, or furniture as well as 2.4GHz.

It's less than two meters. I can actually see the Airport Extreme from where I was sitting. The radio signal does have to go through two panes of glass. I can't imagine that would impair the signal much since it has to travel less than six feet.

(I can imagine some lead-filled glass could cause serious wireless problems :D )
Easy to find out....
Try the iPad in the same room, with only air to the base station.

I'm not about to separate the two networks into separate SSIDs again, but I did check the signal strength comparing the iPad sitting right next to the Airport Extreme and sitting outside with the radio signal passing through two layers of thin, non-leaded window glass.

In both cases the RSSI is −50dBm and the data rate is 65Mb/s. The iPad is connected to the 5GHz network.

On the other hand, my wife's MBA is sitting about six meters from the Airport Extreme on the other side of a brick and concrete wall. The MBA's RSSI is −59dBm with a data rate of 270Mb/s.

An Apple TV3, which is behind two brick and concrete walls has a data rate of only 6Mb/s and a signal strength of −71 dBm.

A brand new Airport Express which is connected via the 802.11a/n network has a data rate of only 6Mb/s and a signal strength of −43dBm. (That's the best signal strength and the worst data rate!) It's sitting right next to an ATV2 which has a data rate of 52Mb/s and a signal strength of −70dBm. This is on the same network.

I could go on....

My network is not dissimilar to your yours. I had problems when I moved into a large 3 bedroom (large concrete apartment building) until I moved things around.

I have an Airport Extreme N Gig Simultaneous Dual Band, Airport Express N, Verizon Router, 2 Macbook Pros, a Macbook, an Apple TV2, Two iPad2, Two iPhones. I too have Extreme Dual Band creating both a 2.4 and 5ghz SSID.

I originally had the Extreme in a centralized part of the the Apartment (wired to Verizon router via apartment wiring) the Express in the Southern most bedroom next to the Apple TV2 and I had the Verizon's Dual band Wifi off. The Verizon Router lives Northern most part of the Apartment in the Network closet. The Extreme is of course Bridged and the Express Extends.

*** My reception was awful. ****

*** I flipped things around and things are great now. *****

I moved the Extreme to the Southern Bedroom (Wired to the Verizon router via Apartments wiring).
I moved the Express to the Central living Room location.
And just in case I needed, I decided to re-activate Verizon's Wifi with own SSID in the most Northern part of the Apartment as it was weekly served by the two Apple routers the at rare times I wanted to use wifi in that part of the apartment.

This change was not intuitive. The Extreme in a central location made more sense. The express should have been more then enough for the Southern bedroom. The Apple TV2 was wired to the network in both layouts. Perhaps having the express and the Apple TV2, 8 inches apart did not help. In my last two homes they were only about 30 inches apart and nothing was wired but the extreme.

The strange thing for me is that the big difference was whether or not I had separate SSIDs for the 2.4 and 5 GHz networks. With separate SSIDs, the performance on the 5GHz network was appalling. Now that I have combined both networks under a single SSID, performance is fine. No video streaming dropouts. No poor iPad performance. It's all good.

I have no logical explanation for this.

5 ghz is great but
 

I too saw the 5ghz issue but untili re-arranged it did not matter which band I used.

5hz (GENERALLY) has superior speed and less interference from other nets
2.4gz (GENERALLY) has greater range and penetration, MUCH more interference

Have you noticed any performance differences between using a single SSID or different SSIDs for each band?

It is an interesting question. In general when the conditions are correct I think separate SSID's have been better. Truthfully I had been under the impression that with the system free to choose it would choose poorly. Though having one SSID I imagine would allow one device to receive the best signal without having to change SSIDs (which usually has to happen manually).

I suppose in theory one SSID should be better if client and router do the best negotiating without human intervention. You can choose which SSID is preferred on the Client side when there is more then one.

It would interesting to see what others have found. My experience has been limited to my most recent 3 homes.

^Which begs the question: Why do you move so often?

I may do some more experimenting when I have more time. For me, the slow-down using two SSIDs was quite dramatic. I set up two because I figured I'd put the devices that needed video streaming on the 5GHz band and put everything else on 2.4GHz. I still think that ought to be a good idea, in spite of what I found in practice.

:) Wife:) We lived in Manhattan for most of our lives. Moved to Florida for 6 years for a lower cost of living, with newborn children in our lives. We moved back for her career and my family 14 months ago. The last move 2 months ago was to a nicer apartment. I hope thats it for a while.

In retrospect looking over my Airport preferences we may have found the same thing. My preferred network preferences were set to 2.4ghz most recently. Essentially removing one of the SSID's from the equation since it is the one with greater penetration. I just collapsed the two SSID back to one. I will report back.

Still getting weird results. The iPad started slowing down again, so I did a Speedtest.net test with a wired MacMini and wireless MBA, iPad (3) and iPhone (4):


While the bandwidth for all four machines is in the same range, the ping times for the two iOS devices are awful. And, this is reflected in the perceived response time when loading web pages, etc.

Why are the ping times so bad for the iOS boxes even though they are on the exact same wireless network as the MBA?

Not sure. Just curious to you use opendns.org for DNS. Highly recommended.

Also I noticed no issues going with the one SSID as expected. You might try rebooting the iPad.

I will try to to run a ping test for you.

I use DNSMasq (running under DD-WRT v24) set up with the following additional parameters:

Any suggestions?

I just ran the speedtest.net for comparison to your results.

Internet Provided by Verizon Fios

Macbook Pro Late 2010 ML versus Ipad 2 IOS 5.1.1 (used the speed test Application
Results were interesting. I did not see the same huge differences though were some.

Macbook Pro
8-9 ms Ping
Download
30Mbps
Upload
22Mbps

IPad2
14-16ms
Downloads
11-15Mbps
Uploads
11-13Mbps

? Though I doubt it is the issue since not all devices are being effected by why use the Buffalo. Have you tried just the Apple Router, also how about just using open dns.org you have multiple dns?

Interesting that your throughput is so much different. Mine was about the same on all devices. Only ping showed marked differences.

I still use the Buffalo out of laziness. I have a large number of fixed IP addresses on the LAN and would have to type all those in to the Airport Extreme. Plus, the Extreme doesn't have enough Ethernet ports, so I'd need a small switch or use the Buffalo as a switch.

Might be interesting as a quick test to use the Apple briefly and to use only one dns like open, but none of these make a logical connection to the Ipad issue. Have you rebooted the ipad and router recently? Why the fixed Ips, Servers?

Not really sure whats going. In my case I am assuming the mobile device is not really capable of using all the bandwidth or gets worse reception? I repeated the test a number of times.

** Update Just found something. The macbook pro (there both in my lap, not at the same time) is a client of the Airport Extreme, the Ipad is a client of the Airport Express.

** I rebooted the Airport Express and this forced the Ipad2 over the extreme.
IPad2
14-15ms
Downloads
27-29Mbps
Uploads
16-20Mbps

I actually started documenting all of the Buffalo's settings so I could migrate to the Extreme. Just one project that stalled out...

I use static leases for the convenience and ease of management. I like to be able to just ssh Bleach without having to know the IP address. Also, it makes port forwarding a lot easier.

I have now set DNSMasq to use only the opendns IP addresses. We'll see if that changes anything.

I still think there has to be some reason why my iOS devices do so poorly on the ping test. In looking at the speedtest.net history I see that my iPad (including my old iPad1) got ping results of less than 100ms until mid June of this year. Since then they've almost always been 500ms+. Same results with the iPhone.

I have no idea what happened in mid June. I've been using the Extreme as a wireless access point since January.

Ha ha ha! The joke's on me. I turned the radio on the Buffalo back on and had the ipad join its wireless network. Here's the results:

ping: 39ms
down: 2.87Mbps
up: 0.81 Mbps

Edit: Loading web pages on the iPad is now astoundingly fast!

So, the fault lies with the Airport Extreme. Amazing.

What now?

Cool. I think you found something similar to I. Were tests conducted in the same spot? I think the iPad may get worse reception?

I have seen a router "mistreat" one device. Often a reboot of both helps.

All interesting.

Yes, interesting.

I changed the wireless security setting on the Extreme from WPA/WPA2 Personal to just WPA2 Personal. Now the ping time on the iPhone is in line with everything else; about 35ms. However, the iPad is still slow at over 600ms; even after a reboot of both the iPad and Extreme. Note that the iPad is connecting to the Extreme's 5GHz network.

So, I split the Extreme's SSIDs and forced the iPad to join the 2.4GHz network. Still slow.

Then I forced the iPad to join the 5GHz network. Now the ping times are less than 40ms.

Bottom line: I don't understand this at all.

Will a wireless expert please chime in?