Q: Intermittent Wifi issues with iPad 2 and WiFi

JimHdk wrote:
That feature is an optional part of the 802.11n standard. It is not required for certification. Use of that feature is not advised in the 2.4GHz band since its use can create interference chaos in an already highly congested frequency.

Well aware that many people turn it off for the reasons both you and I stated. The Wikipedia article you reference certainly doesn't say it's an optional part of the standard for certification so you must be getting that info from another source?

I came across the following advice from Microsoft's xbox.com website:

"Tip 3: If your router is set to "Auto", "20/40MHz", or "40MHz" and using 2.4GHz, try setting it to 20MHz only. You will probably have higher bandwidth, lower latency, and less disconnects due to interference."
http://www.xbox.com:80/en-US/Live/EngineeringBlog/043010-WirelessNetworking

So it seems that Apple is not the only one to think that use of the 40 MHz band can result in disconnects.

My fix and setup:
DIR-655 wireless N router
Mix b/g/n
Channel 11
WPA2 only
AES encryption

Has been solid for a while. Fixed the constant reconnect prompts I was getting.

tgibbs wrote:
"Tip 3: If your router is set to "Auto", "20/40MHz", or "40MHz" and using 2.4GHz, try setting it to 20MHz only. You will probably have higher bandwidth, lower latency, and less disconnects due to interference."
http://www.xbox.com:80/en-US/Live/EngineeringBlog/043010-WirelessNetworking

So it seems that Apple is not the only one to think that use of the 40 MHz band can result in disconnects.

Using 20MHZ instead of 40MHz gives you half the bandwidth and less than half the throughput (less than half due to overheads) in an environment without interference. The reality is that in non greenfield installations using 40MHz makes you much more susceptible to interference. In installations without interference options it 's nice however to have over twice the data throughput.

FastBikeGear wrote:

JimHdk wrote:
That feature is an optional part of the 802.11n standard. It is not required for certification. Use of that feature is not advised in the 2.4GHz band since its use can create interference chaos in an already highly congested frequency.

Well aware that many people turn it off for the reasons both you and I stated. The Wikipedia article you reference certainly doesn't say it's an optional part of the standard for certification so you must be getting that info from another source?

The article does state that 40 MHz is an option in the standard. That option is not mandatory as per the IEEE 802.11n specification itself. There are a lot of other optional features in the standard as well. Apple devices are certified as 802.11n compliant.

Many were against including 40 MHz at 2.4 GHz in the specification at all due to the interference problems. Apple chose not to implement 40 MHz at 2.4 GHz for this reason.

By the way, all this has nothing to do with the WiFi problems posters are having.

JimHdk wrote:

The article does state that 40 MHz is an option in the standard. That option is not mandatory as per the IEEE 802.11n specification itself. There are a lot of other optional features in the standard as well. Apple devices are certified as 802.11n compliant.

By the way, all this has nothing to do with the WiFi problems posters are having.

The article you are referring to is another great Wikipedia article. The article states that being able to select between 20MHz and 40 MHz operational is optional (and it is correct to state this) but it does not state, that not being able to support 40MHz is an option for Wi-Fi devices. This is a fine distinction.

Nor does the article you quote say that having the devices not declare their 20/40MHz mode of option is optional.

Reading the full source of the article may make things clearer.

From my limited familiarity with the standard (It's been several years since I originally waded through it - and I had to get it out this morning to scan through it again) it seems that the IEEE 802.11n standard does require both AP and STA's to 'handshake 20 or 40 MHz operation".

Check out section 11.14.1 & 2 of the standard where it talks about both the AP and the STA declaring channel width capability.

Signalling of 20 or 40MHz option does not seem to be optional. It seems to be a necessary requirement. Signalling of 20MHz or 20MHz mode is done by setting bits in the Supported Channel Width Set sub field.

Apple has stated they don't support the 40MHz mode in the iPAD. OK we can live with that (although ideally the members of both the Wi-Fi Alliance and the IEEE agreed and voted to have this option so that we could get more than twice the bandwidth in environments where interference is not an issue.)

But the iPAD is required to signal whether it's operating in 20 or 40MHz as part of the 'hand shaking" that goes on between Access Points and Stations. Therefore it should not be a requirement to have to manually disable the 40MHz mode in the router when a Wi-Fi compliant device that is set to operate in 20MHz mode connects with it.

Further, while in the standard it talks about the options of being able to turn off the 40MHz mode, nowhere that I have read in the standard (so far) does it say that not being able to support the 40MHz mode is an option of for 802.11n compliant devices.

We know from Apple that a number of their 'Wi-Fi' devices do not support 802.n 40MHz channel. Because they ask you to manually turn off the 40MHz mode in the Access Point as a work around, we can suspect that they may do not handshake/negotiate as per the requirements of the standard.

However given the Apple iPAD limitations, manually turning off the 40MHz mode in the Access Point as a work around seems like good advice and by logical extension is one of the reasons why I believe that some people have had success by turning off 'n' mode completely in their routers.

I am becoming more and more convinced the issues are in the way the iPADs handshake with the Access Points.

Message was edited by: FastBikeGear

A quick insight into what is mandatory and optional in the IEEE 802.11n standard is given by the following table:



Since the iPad implementation of IEEE 802.11n (via Broadcom BCM 4329) is certified it's safe to assume that the mandatory elements are present.

That's more than enough on this subject i'm sure.

I have the same issue with my 32G Wifi iPad 2. I have a Linksys E3000 router, I've tried various modes however the problem persists. In my case the connection is stable during a session until I close the smart cover and the device hibernates, 9 times out of 10 when I start up again the Wifi fails.. If I hit the renew lease in the wifi settings the connection is restored.

JimHdk wrote:

Thanks for the table. I don't recall seeing the Table in the 802.11n Standard? I am guessing it is someones interpretation derived from the Standard's provision for turning off the 40Mhz operation in sites where interference is a potential issue.

I did a quick, but not exhaustive look through my copy of the IEEE 2009 802.11n standard (It's 496 pages long) I searched on the word 'optional' (which is used frequently in the standard), and I also glanced through the sections that I thought might also be relevant to whether 40MHz operation is mandatory.

Certainly their are many features specifically listed in the standard that are noted as optional. But I didn't see anywhere where the Standard states that the 40MHz capability is optional. I admit that I have assumed that if a feature was not noted as being optional that it would then be mandatory. Maybe that's not a reasonable assumption. If anyone cares about this issue they can read the standard and make up their own mind on this issue. You should be able to find and download a copy of it on the the www.wi-fi.org site

Regardless, I think the need to be able to turn off the 40MHz option for interference issues is well understood and the standard certainly makes provision for turning this off and declaring this in the 'handshaking'. Certainly within the standard it would seem that there is a requirement for the STA and AP to 'declare' if they have 40MHz operation switched on or off.

It would of course be nice if the iPAD did support the faster data rates specified in the 802.11n standard.

I think the key issue that could impact on the intermittent connection issues is not that Apple say that the iPAD does not support the 40MHz operation detailed in the standard but whether or not the iPAD is correctly declaring to the Access Point that it does not support this mode during authentication and re-authentication bit sequences. (Authentication and re-authentication are slightly different processes and I think the issue is related to the re-authentication).

JimHdk wrote:
Since the iPad implementation of IEEE 802.11n (via Broadcom BCM 4329) is certified it's safe to assume that the mandatory elements are present.

Careful JimHdk, shortly you will convince us that this problem is a figment of our imagination . I get the feeling that your understanding is the intermittent connection issue are being caused by the routers? (if I have made an incorrect assumption here, let us know where you think the problem lies or whether you think it even exists.)

I am not so sure that saying the chipset is made by Broadcom carries much weight. I've been involved in computer network engineering for long enough to know that the brand of vendor is no guarantee of a perfect product and we do know that Apple (no more than other vendors) has had product issues in the past.

Your assumption however that their is no problem with the iPad WiFi connectivity, because it is certified is a good one.

It should also be noted that I can reproduce the problem with a Wi-Fi certified router from Cisco (WAG120N).


It might also be relevant that ony the iPAD with the first version of the iOS was certified. It has not been recertified with subsequent iOS updates. As stated previously by myself and several other posters on this thread we had no problems with intermittent WiFi connection with the first iOS. However it might be possible to try and determine too much into this observation.

If someone can advise me how to embed a picture in posts in this forum I will paste some stuff on Authentication and re-authentication from the standard that might be technically relevant to the problem people are imagining they have.

Message was edited by: FastBikeGear

JimHdk wrote:
Since the iPad implementation of IEEE 802.11n (via Broadcom BCM 4329) is certified it's safe to assume that the mandatory elements are present.

Actually while the iPAD with the first iOS was Wi-FI certified the Broadcom BCM 4329 appears to have never been WiFi certified. I am not even sure that the WIFi alliance certifies chips so I am not sure what "via Broadcom BCM 4329" refers to.

The more you dig, the more interesting stuff you find. The Broadcom BCM4329 chip used in the iPads was first released in Dec 2008 before the 802.11n standard was finalised/ratified.

From Broadcoms release notes the chip is a dual band (2.4Ghz/5Ghz) wireless chip but it is only a single stream radio device! It does not support MIMO, it has a maximum data rate of just 50Mbs and while it supports a subset of the features in the 802.11n standard I can't see where Broadcom ever claimed it was WiFi 802.11n certified.

The advantage of only having a single stream and consequently being a non MIMO capable radio device is that it consumes less power. The disadvantage it that while their are some performance gains over 802.11g and 802.11a they aren't very significant. 50Mbs throughput is probably good enough for most things you want to do on an iPAD though.

This info may be a bit of a red herring because many users in this thread and threads on the same issue in other forums state that their other Apple devices do not have the same intermittent wireless connection issues as their iPADs when working on the same Wireless Access points and routers. Some Apple devices use the BCM43224 chip set.

It also has on-board power management but this does not necessarily prevent Apple implementing further more aggressive'parental' power management control over operation of the wireless in the iPAD.

Broadcoms BCM43224 chip does support 2 radio streams is MIMO capable and is WiFi certified...but no doubt consumes more power. This chip is used in their WiF-Fi certified reference design products. e.g. http://www.broadcom.com/products/Bluetooth/Bluetooth-RF-Silicon-and-Software-Sol utions/BCM943224HMB

ref: http://www.wi-fi.org/searchproducts.php?search=1&lang=en&filter_categoryid=4&listmode=1


ref:http://www.broadcom.com/press/release.php?id=1233460

There are quite a few other resources on Broadcom's very informative web site for this chipset

Message was edited by: FastBikeGear

JimHdk wrote:
Since the iPad implementation of IEEE 802.11n (via Broadcom BCM 4329) is certified it's safe to assume that the mandatory elements are present.

Actually I just checked the iPADs certification certificate. It is only 802.11n draft 2.0 certified! To put this in perspective the drafts went all the way up to draft 11 with many significant revisions along the way until the standard was finalised and published. The reason I double checked was because something smelled very fishy. I couldn't see how an '802.11n' device that used the Broadcom 4329 chipset could possibly be certified.

I had assumed previously that it had achieved 802.11n certification simply for the reason that it's certification testing was done in 2010 and the final standard was ratified in 2009. Given the dates you have to ask why it wasn't certified to the ratified standard and only achieved certification to a draft standard that was years old.

For Apple to claim the iPad is an 802.11n device seems a bit misleading!

ref http://www.wi-fi.org/searchproducts.php?search=1&lang=en&filter_categoryid=24&listmode=1

The iPAD has never been Wi-Fi certified and never has any other product using the Broadcom BCM 4329 chip that I can discover so far.

Message was edited by: FastBikeGear

Mark me down as another example of this problem. I try to show off my 64 GB Wifi iPad 2, but it gets embarrassing when I have to reset the wifi about every 10 minutes.

My iPhone 3GS and iPod Touch 4th Gen have no problems maintaining connects on the same wifi at the same time.

Same wifi problem. Rebooting solves problem for a short time

Interesting? I've only read back to page 10 of this thread looking for references to others setting a static IP and whether or not they saw any improvement (i didnt see any)I read in an earlier post the user wondering if re-authentication may be the problem and another wondering about DNS settings. This clearly doesn't help when out looking for hot spots but maybe provides bread crumbs? I set a static IP for my iPad last night and have not had a single instance of failure today.

Jtatton wrote:
Interesting? I've only read back to page 10 of this thread looking for references to others setting a static IP and whether or not they saw any improvement (i didnt see any)I read in an earlier post the user wondering if re-authentication may be the problem and another wondering about DNS settings. This clearly doesn't help when out looking for hot spots but maybe provides bread crumbs? I set a static IP for my iPad last night and have not had a single instance of failure today.

I tried a static DNS setting as that has reportedly worked for other users. For me it didn't work. Can you update us in a few days and let us know if it is still working for you? Every bit of knowledge helps the rest of us.