• Hi folks.

    I am trying to find the table of data rates that the most common apple devices support and the signal strenght at which they change data rates.

    If anyone has any info please would you let me know (and it's a bit urgent...)



  • Disclaimer: I'm no expert on apple products...

    It's my understanding the that ipad and ipad2 have a dual band 11n radio with support only for a single spatial stream using the BCM4329 chipset with a maximum theoretical throughput of 65Mbps

    The speed at which the device negotiates is dependent on both the device itself and the wireless station that it's associating to.

    This table may provide some insight into what data rates you may see,

    1 spatial stream, 20Mhz channel width, long guard interval (800 ns)

  • Thanks, Ryan, I know that already though.

    The WiFi alliance certs for each device add to what you said and have more detail.

    I am fine with the theory of the rates, I am trying to find specifically the dbm values at which the chipset (as installed in the iphone) is set to shift.

    I want to see what rater will be used at -70dbm, -75 dbm, etc.


  • I am betting that the only way you will find that information is to test it for yourself - although if you were a big enough Apple customer they may give it to you (NDA ?).

    Rate adaptation is a much more sophisticated process than it used to be. We often hear that roaming algorithms are part of the "secret sauce" for individual radio (i.e. not just chipset) manufacturers. Rate adaptation is a "kissing cousin" to those algorithims. These may be tied very closely, and be difficult to separate.

    While trying to determine the "hard and fast" adaptation points may not be too difficult, it might be difficult to include retry and missed beacon counts into your measurements.

    It will also depend on the accuracy you desire. Are you looking for dBm ratings of the same quality that Apple uses to define their performance ? Something corresponding to the values that your site survey equipment provides ? Or maybe just InSSIDer ?

    If you've got the big bucks, go with something from ETS-Lindgren. They have prepackaged systems that are alreay CTIA and WFA certified.

    If you can find someone with Veriwave equipment, you may be able to get them to run a test for you too.

  • Nothing so complex.

    if you look at the spec sheets for a lot of wifi devices they will give you the modulation rate at specific signal strengths.

    I only want to see the ones for apple devices.

    I would like to know what modulation is used at around -70dbm.

    It should be simple enough.


  • I was trying to put things in general terms for testing anyones gear, and I don't think that is quite as easy as you think it is.

    [i]I think your absolute best bet is to find someone who either already knows for sure, or has a good rule-of-thumb for the Apple devices that they support[/i].


    Back to the general case - The Chipset is probably from one of five, or so, different companies. But a radio could be from a hundred different sources. On top of that, each device manufacturer MAY if they wish (and want to pay the radio mfg. for the privelage) modify the roaming and rate adaptation mechanism to meet their own needs. [b] It's a sure bet that these parameters are definitely part of Apples "secret sauce"[/b].

    The numbers specified by any radio manufacturer will be for [u]conducted measurements[/u] as per the IEEE specification. Do you have a way to do that kind of test? If not, you will only get a rough estimate - in which case you'll have to do the exact same kind of test against whatever you are comparing it against.

    The specifications given for "[i]the modulation rate at specific signal strengths[/i]" may have nothing to do with the same devices rate adaptation algorithm. From a compliance point of view, the values are probably given as [i]signal strength at specific modulation rates[/i], anyway.

    On top of that, manufacturers usually generalize about performance. Some of which is justifiable, and in other cases, at best, misleading. I've run thousands of tests on different radios and devices using high precision laboratory grade equipment, so I'm not just making this up.


    BUT, I forgot to add yesterday, that if you want to make the best measurements possible:

    - Disable Roaming, and/or

    - Disable Power Save on the client

    Good luck !

    PS: Like you, I am hoping someone can provide a definitive answer to your question. I always like to hear about a manufaturers designs from another device domain.

  • Thanks for the answer, however I think you are going too deep.

    I am not going to test this, I am planning a deployment.

    We can't upload pictures in posts here (pity) but go here:

    Scroll down to "Receive Sensitivity" and you will see a table of rates for power.

    All I am looking for is thge iphone equivalent.

    Who decides the data rate ? If it's the AP then all well and good, it will be determined by the AP RSSI (discounting other factors, I can go with the above table for this example).

    Is it the iphone ? In which case, where is the spec sheet as above ?

    I understand that there will be additional factors, I don't care about those for now (interference, noise, etc), it's not relevant.

    I just want to know, under ideal conditions, what modulation will an iphine try to use under best conditions at -70 dbm, then at -75dbm.


  • I have found that Cisco specifications, especially for sensitivity are some of the best, and the most honest numbers advertised. Thankfully.

    Up to this point, it has sounded like you wanted rate adaptation points, not receive sensitivity values. They are not the same thing at all.

    Sensitivity, has no [i]fixed[/i] relation to a rate adaptation algorithm, [u]except[/u] that the goal will be (today) to shift the rate downwards before the minimum received signal is passed. This keeps the link from being lost completely if the client does not find a better signal (AP) quickly.

    As others will atest, we have found that sensitivity is a much better predictor of maximum range than output power. This is provided we have some semblance balanced power between the AP and the client.

    If all you wanted to do was determine the best possible value for a particular device's sensitivity (without any test equipment), you could determine(?) the radio used, and find what the RADIO manufacturer's sensitivity claims are.

    A manufacturers numbers will (again) be based on conducted, not radiated measurements - meaning no antennas. The (radiated) sensitivity will be based upon several factors, too many to go into in this thread.

  • [quote]Who decides the data rate ? If it's the AP then all well and good, it will be determined by the AP RSSI (discounting other factors, I can go with the above table for this example).[/quote]

    When a device/spec sheet describes rates in relation to RSSI refers to the TX rate for that device. The AP and STA each choose a TX rate independently, and they rarely end up being the same. Even if they are the same, they won't stay that way for long -- that's the [i]dynamic[/i] in dynamic rate switching :-)

  • We recently deployed some products manufactured by Apple in our environment. The iPhone as aWiFi client tends to be rather "sticky" along with some other "peculiar" behavior. Remember this is a consumer grade device so when it lataches on to an "AP" it really doesn't want to let go.
    I could not get any info from Apple regarding the receive sensistivity and rate change. So... I spent a lot of time analyzing AirMagnet captures and listening to the users about how these devices behaved as well as testing we did in our IS shop. This is what worked for us.
    Note: We are using these devices for both voice and data and we are a Cisco shop. We are running RRM but I found that the minimum power levels had to be set 3db hotter than what Cisco's standard best practices say. In addition, we set the minimum data rate to 12Mb. So far this has worked to improve roaming handoff and voice quality and we have had no adverse impact on the other 17 clients we have to provide WiFi service for. Hope this helps...

Page 1 of 2