Re: [ns] RTS/CTS Handshaking in 802.11

[Anders Lindgren <dugdale@sm.luth.se>]

Hi,

If you look inside the function Mac802_11::sendRTS(int dst) you will see
the following lines of code:

  if( (u_int32_t) HDR_CMN(pktTx_)->size() < macmib_->RTSThreshold ||
      (u_int32_t) dst == MAC_BROADCAST) {
    Packet::free(p);
    //p = 0;
    return;
  }

Meaning that is the size of the packet to be sent is less than the
RTSThreshold (or if the destination is broadcast), RTS/CTS should not be
used, so the packet is deallocated, and the function returns without doing
anything else. 
It does however sound strange that it doesn't affect the performance in
your simulations (low load could lead to no change in throughput, but as
you say the delay should be higher...). What metrics are you looking at,
and how do you measure them?

/Anders

On Wed, 31 Jan 2001, Al-kahtani wrote:

> Hi Everyone, In IEEE802.11, I tried to change RTSThresholed value
> (mac-802_11.h) to see RTS/CTS handshaking effect on the performance.
> My packet size was 512 bytes and the RTSThresholed default value was
> 3000. When I change it to 300, nothing has changed in the performance
> and as u already know it will affect at least the delay of the packet.
> When I go through mac-502_11.cc source code I found RTS/CTS is used
> regardless the RTSThresholed, eventhough this idea is mentioned in the
> comment .  Am I right?. Have anyone checked that before?
> 
> 
> Regards,
> Mohammed Al-kahtani
> 
>