Wireless Access

last person joined: 16 hours ago 

Access network design for branch, remote, outdoor, and campus locations with HPE Aruba Networking access points and mobility controllers.
Back to discussions
Expand all | Collapse all

Testing Air time

This thread has been viewed 1 times

  • 1.  Testing Air time

    Posted Mar 09, 2012 02:39 PM

    Does anyone  knows any easy program or tool to test the air time? with graphics to see before and after airtime configuration?

    i would like to see it working not just configure it i would like to know its working as it should.

     

    I mean on aruba network book there is a example of a graphic how was it before air time and after and i would like to do the same...

     

    Anyone?



  • 2.  RE: Testing Air time
    Best Answer

    Posted Mar 09, 2012 03:50 PM

    If you are talking about air time fairness a easy way to test is using tools like Chariot or the free iperf. Have a mix of a, g , and n clients. Now run high UDP or TCP traffic to each client and see the performance of airtime fairness by noting the thorughput achieved for each client. Run the test for all airtime fairness modes (default, fair and preferred access mode) and see how the thorughput to the a/g and n clients vary base on the airtime fairness mode. When you are testing in the 2.4GHz band have atleast one or two g clinet and two n clients. Similarly for 5GHz have atleast one or two a clinet and two n clients.

     

    Note: It is best practice to test airtime fairness with UDP traffic type so that the test purely measures the AP's ability to shape traffic without taking other factors into account like TCP windowing, congestion control, etc. If the Fairness test cases use TCP traffic type the results might not be very accurate due to TCP parameters, which does not necessarily mean that the airtime fairness is not effective.

     

    Regards,

    Sathya



  • 3.  RE: Testing Air time

    Posted Mar 09, 2012 04:12 PM

    Thanks man, ill get those tools

    Any recomened number of computers? do i need like 30 computer or something like that?

    Because to be honest i got like maybe 5 or 6 computers avalaible for the test, it would be a bit troublehsome getting more, but if its needed i guess ill ge them.



  • 4.  RE: Testing Air time

    Posted Mar 09, 2012 04:55 PM

    That is good enough. You don't need 30 computers to see how airtime fairness works. The important thing is to have  a couple clients as non-HT clients i.e. as only a or g clients. On windows machine you can easily do this (it may vary for diferrent wireless cards, this is for intel)

    wireless nic.png

     

    wireless nic 2.png

     

     

     

     




  • 5.  RE: Testing Air time

    Posted Mar 09, 2012 06:11 PM

    sathya thank you very much...

    It jus that i plan to create a lab in which i have all this set up so i can show the clients the different aspect of aruba and why they should buy aruba :)

    And well this is something i wanted to show when im showing them a demo.

     

    Kudos!



  • 6.  RE: Testing Air time

    Posted Mar 20, 2012 05:41 AM

    Why use A & G clients?  To test airtime, don't you want to see the max throughput that a transport can provide? The distinction would be which path can provide the fastest throughput.



  • 7.  RE: Testing Air time

    EMPLOYEE
    Posted Mar 20, 2012 06:18 AM
    @dfc wrote:

    Why use A & G clients?  To test airtime, don't you want to see the max throughput that a transport can provide? The distinction would be which path can provide the fastest throughput.

    I think it was mentioned to ADD A and G clients to the mix that would already have 802.11n (HT) clients...