Forum

  • Just a quick problem that I am trying to get my head around and hopefully someone might be able to answer or point me in the right direction.

    So, after discovery/authentication/association ... the STA and AP are communicating.

    I know that within the beacon frame, basic rates etcwere defined.

    Page 301 CWNA 106 covers this and says all basic rates must be supported.

    So, how is the actual rate decided ? I assume SNR/RSSI/Quality etc.

    Will the data rate be the same in both directions?

    I.e can or would the AP transmit at a higher rate than the STA could?

    Thanks

    Chris

  • By Howard - edited: October 23, 2015

    Chris,

    Typically AP's transmit with more power than client devices.   One reason for this is that most clients run on battery power, and expecting users to be changing batteries all the time is not realistic.   Most handheld barcode scanners, printers, etc. are designed to run for at least one shift,  or even 10 to 12 hours, between changing or charging batteries.

    There is also the concept of "balanced" power levels, but I'm trying to keep this explanation short.

    Client devices usually have one or more power save algorithms to save power, and some even shut off completley if not used for several minutes.     The expensive gear usually has tunable settings for controlling such behavior, but these may be set by a system integrator, rather than the final user and the customer may not even know such settings exist.  I am talking about enterprise quality gear now, not necessarily household devices.

    Some settings don't really do much, but they make the users feel more in control.  Others have large effects, but take considerable tuning effort

    Basic rates are set on the AP, by the system administrator or integrator.   They are ususally set based on performance and range requirements, with a good dose of "experience" rolled in.   Often the AP or Client manufacturers will make their own recommendations as to what should be used..

    The actual rates selected depend on the rate adaptation algorithm(s) in each device.   These are the manufacturer's "secret sauce", and are usually very closely held secrets.  Roaming, authentication, power saving, channel selection, active or passive probing, etc - all of these are used in the special sauce.

    You would think that SNR would be an obvoious parameter upon which to make roaming and rate adaptation changes, but often it's simpler to just use missed-beacons and retry counts to drive the process.   Power levels will be in there somehwere, but not as likely as you might expect.

    One other factor to consider is that different site environments may also affect performance.  A large warehouse or office are very different, and one manufacturers equipment may work better in your space than an other one using the same chipset.   They tune them differently based on their experience.

    It's all in the "secret sauce". 

  • Thanks Howard,  makes sense when you put it like that.

    I suppose that both the AP and STA would have to have a degree of independence based upon proprietary algorithms etc...after all there needs to be flexibility when you consider the number of differing devices out there.

    thanks for the great explanation.

    Chris

  • Hi,

    I passsed CWNA exam last week and although I had quite a solid result (90 %) I think I made a mistake when answering to one question concerning speeds which are used by the AP and client station to transmit.

    I thought that both of them agree to use the same speed during the discovery, authentication and association process but later I have discovered that it seems that both of them are using different speeds that are set according to their experience of RF channel.

    Please, correct me if I am wrong.

    Tarik

  • Congratulations Tarik,

    I think Howard's explanation probably answers the question for you.

    I had it in my head that both the STA and the AP would connect at the same "Basic Rate", but if you think about it that is not happening due to the factors Howard mentions... and add DRS into the equation and it complicates data rates between the STA and AP even more.

    I have my CWNA exam on Friday 10:00 ... I cannot say I'm looking forward to it :-)

    My "day job" is Computer Forensics, and although I am not looking to do wireless administration I do want to get deeper into wireless security... so CWSP would be my next exam ... assuming all goes well!!!

    Regards

    Chris

  • There is no single, simple, answer to this question.

    An early approach was to reply to a message at the same rate at which it was received, but this turned out to not work under all conditions.   Hence different rate adaptation algorithms were designed by different manufacturers.  If you read through a wireless packet trace you'll often see these differing rates being used.

    Remember that there are several categories of rates.   The ones most often listed are Basic (aka Required), and Supported Rates.  By important implication then, there are also Unsupported rates.  Some manufacturers refer to these categories differently, and the settings on some SOHO AP's may only let you set a Maximum rate. 

    Each side of a connection may have its own unique preferences, whch further complicates the situation.   These may be set by a network admin, an integrator/distributor, the user, or the original device manufacturer.

    The simplest situation is where the client supports all of the rates, and the AP's settings are used to tune the overall network performance.

    From an Association point of view, the Basic rates are the final arbiter of whether a client can first connect to an AP.   After that point, the individual rate adaptation algorithms on each side of the connection make the selection of what rate communications proceed at.   Again, this is part of each sides "secret sauce".

    Because clients and AP's often have different Sensitivity and Output Power levels they may make different rate selections even if they use the same adaptation algorithm.

    As AP's are more expensive, they usually have better Sensitivity, and almost always have more Output Power potential.   Being continuously powered they also do not have to use, sometimes overly complicated, power-save mechanisms.   WMM-PS is an attempt to overcome this latter issue, but not all devices support it.  Recent Energy Star related changes seriously confuse this issue, on any mains powered STA, as well.

    One side effect of all this is that just because a device can receive at a high rate, doesn't mean it ever transmits at that rate.

    Believe it or not, some parts of the 802.11 statndard, which may seem very clear and direct, are interpreted differently and/or more loosely by one manufacturer than another.   Take for example the requirement to always support the 6, 12, and 24 Mbps rates of a/g.   Does this mean the "hardware" must support it, or the firmware, or the configuration in the device?   Believe me, you will see differences.

    When in doubt, trust your packet analyser. 

  • @Chris

    Thank you, I hope that you'll pass the exam on friday.

    Great help for me was the assessment test that is in the book and all the other tests which are provided by the CWNP in the study materials. You can also find some demo versions of the CWNA exam in the Internet. By taking these tests you switch to the exam mode. I would suggest to you to prepare yourself in that way on thursday and to relax as much as possible on friday without taking any further exam preparations. If possible of course.

    @Howard

    Thanks on this thorough explanation

  • Thank you Tarik,

    I have used the CWTS, CWNA and CWSP books from Sybex and also the CWNA e-Learning course from CWNP.

    The books are really good and go into good detail, maybe more than needed for some areas???... but they will also be  great reference material.

    The e-Learning is also good value as it gives 2 practice test pools along with "mini-tests" for each chapter...AND an exam voucher.

    The instructor obviously knows his stuff!!! ... but  wish he would SLOW DOWN a little... you do not have to speak at 802.11ac rates :-)

    So tonight is really revision ...hopefully an early finish from work, a couple of hours DE-ciphering my pages of notes and an early night!

    Cheers

    Chris

  • Well I managed a 93% on the test so I was over the moon!

    I have to say that the on-line practice tests and the e-Learning course, in conjunction with the Sybex book and recommended white papers  and people on here (Howard) helped tremendously.

    The practice tests and the QOTD are very close to the live exam, more so than the Sybex, but having said that the Sybex tests do cover the objectives in great detail.

    I guess it is now onto the CWSP, I'm having a little break first then aiming for end of the 1st 1/4 next year.

    Chris

  • Congratulations.

    Great score     !!!!

Page 1 of 2