> 
> RTS/CTS is a good feature in case where clients can
> not hear each other.
 In case of Infrastructure mode too, 2 client nodes
can be hidden from each other.
> RTS/CTS has an added overhead of sending RTS/CTS and
> then real data can
> be transferred. This needs to be performed for each
> packet transmitted.
  Isnt it Worth the overhead to prevent data
collisions. Data collisions are costly in the sense
much more slots are wasted than during a RTS
collision.

> If packet size is small, RTS/CTS could add high
> overhead and hence
> reduced throughput. Performed tests and network
> simulation shows that
> RTS/CTS may be good in case of large packet size.
>
   RIght , thats why in my opinion, that RTS threshold
should be set to an optimum value (some papers say
arnd 200)
 
Harish
> Hope this helps
> Lalit
> 
> 
> 
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of
> balan h
> Sent: Saturday, February 21, 2004 1:00 AM
> To: [EMAIL PROTECTED]
> Subject: [BAWUG] RTS threshold question
> 
> Hi,
>  The RTS threshold value in many Access ports is set
> to 2300+ bytes as default value. This is apparently
> to
> disable RTS/CTS exchanges. I dont understand why we
> do
> this.
>  Isnt the RTS/CTS exchange used only in the
> 'Contention Period' of 802.11 where the Access point
> does no polling? Then shouldnt we enable RTS/CTS
> exchange most of the time by setting RTS threshold
> to
> a low value, to avoid collisions?
> 
> 
> balanh
> 
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail SpamGuard - Read only the mail you want.
> http://antispam.yahoo.com/tools
> --
> general wireless list, a bawug thing
> <http://www.bawug.org/>
> [un]subscribe:
> http://lists.bawug.org/mailman/listinfo/wireless
> 


__________________________________
Do you Yahoo!?
Yahoo! Mail SpamGuard - Read only the mail you want.
http://antispam.yahoo.com/tools
--
general wireless list, a bawug thing <http://www.bawug.org/>
[un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless

Reply via email to