Forum

RTS/CTS question

20 posts by 5 authors in: Forums > CWNA - Enterprise Wi-Fi Admin
Last Post: June 9, 2005:
  • Hi

    I am trying to understand the RTS/CTS process more clearly and the description of the process in the Study Guide V2, p.340 is somewhat vague.
    Basically, the book says that the sending station sends an RTS frame to the 'intended recipient'.
    Since all traffic between stations in a BSS must flow through the Access Point, wouldn't this be a 2-step process?
    By this I mean, wouldn't the first RTS frame be sent to the AP, with the AP returning a CTS to clear the medium, then receive the data packets, turn around and send its own RTS to the recipient, wait for the CTS and finally deliver the data to the destination?
    Thanks for the help.

    Denis

  • Hi dharrison,

    As I understand your own answer, to your ?,, yes. RTS/CTS is normally used as a protection. The AP is the one that tells all to use it. Its a way to reserve the media. So yes. If laptop A, is talking to Laptop B - through AP 1. Then its a two part. The AP will need to tell all the laptops to be quiet with a RTS. As they are all listoning to the media, they will be quiet untill the "time that was posted in the RTS frame header has passed"
    BUt if your talking to the DS (distribution system, normal ether) than no. Or at least not in that BSS.

    Well, that is what I belive is happening. I googled it. No tmuch better...

    Phil

  • Hi Denis of Burnaby:

    You are correct, and the paragraph is vague.

    "Intended recipient" is the immediate link partner whose MAC address is found in the "Address 1" field of the MAC header.

    The explanation in CWNA study guide v3 is better.

    I hope this helps. Thanks. /criss

  • Hi Criss

    As usual, your reply is almost instant. That's what I call quality of service!
    Further to the question I posed, the text also implies that reason the other stations remain 'quiet' is because they hear the CTS frame. Is this really correct?
    If RTS/CTS is being employed, it has to be implemented on every device on the WLAN, doesn't it?
    If that is true, wouldn't all devices be quiet by default unless a (unicast?) RTS frame arrives at a particular node, which can then and only then be allowed to return a CTS and then receive data?
    Your help will be appreciated.

    Denis

  • Hi Phil

    Sorry for not thanking you for your reply, but I was responding to Criss' e-mail.
    I noticed your response when I went back to the forum.
    Thanks again for yur help. I will try to think about all this for a while.

    Denis

  • Hi Denis of Burnaby:

    Assume a single infrastructure BSS with many stations all within range of the access point (AP) but not all within range of each other. Client1 is out of range of client2; each is a so called "hidden node" (an unfortunate term coined by the IEEE 802.11 Handbook) regarding the other. Notice that a "hidden node" never exists in isolation but rather only in pairs.

    Assume client1 has a frame to send to AP. Client1 sends RTS. Clients hearing RTS set their NAVs and defer. AP sends CTS. Clients hearing CTS, including clients like client2 that did not hear the RTS, set their NAVs and defer. Classic.

    RTS/CTS is configured station by station by an administrator. With a single BSS only the "hidden nodes" as defined above would benefit from using RTS/CTS and these would never include the access point. However multiple BSSs occupying or overlapping the same channel create a much more complex interaction in which even the access points may benefit from using RTS/CTS.

    I hope this helps. Thanks. /criss

  • No Problem!! I'm new at all this to. I "just" passed the CWNA. I try to help out as freely as it was given to me. Plus, it helps me lurn it even better. As you can see, I'm a bit long winded to get the point across :).
    Have a good one!!

    Phil

  • Thank you Criss and Phil for your replies. I am starting to understand this better now, although the last paragraph in Criss' response got me confused.
    In it he states "only the hidden nodes would benefit from RTS/CTS and these would never include the access point..."
    I thought no communication could take place within a BSS without the participation of the access point.
    Sorry to be such a pest. In the future, I'll only post when I am really, really stuck.
    Thanks,

    Denis

  • Hi Denis of Burnaby:

    With a single BSS only the "hidden nodes" as defined above would benefit from using RTS/CTS and these would never include the access point. Every client station in an infrastructure BSS is within range of the AP, or the client loses association and drops out of the BSS.

    Assume mutually hidden (out of range of one another) client1 and client2 are configured to use RTS/CTS, the AP is not configured to use RTS/CTS, and client1 pings client2 with a relatively large ping packet.

    Client1 sends the ping request payload to the AP preceded by RTS/CTS and gets an ACK from the AP, the AP sends the payload to client2 without RTS/CTS and gets an ACK from client2, client2 sends a ping reply payload to the AP preceded by RTS/CTS and gets an ACK from the AP, and the AP sends the payload to client1 without RTS/CTS and gets an ACK from client1.

    I hope this helps. Thanks. /criss

  • In Criss's example, he is stating that the AP is not using RTS/CTS. It is only answering the Laptops RTS. And in answering with a CTS the "hidden node" now hears the fact that someone is sending and to not send.
    It's like have a mediator. He is heard by all partys!
    And the people to far away to hear the others, respond to his "ok, say it" and everyone will lisson.
    Until the mediator sends an ACK. BUt the Mediator is not using RTS/CTS. He is only answering the people out there..

    Keep asking! But it might help to get an understanding of the Hidden node. And why this will help them. It will all come in time.

    Keep doing it!!

    Phil

Page 1 of 2