Energy and GTS

46 views
Skip to first unread message

sara so

unread,
Nov 9, 2009, 6:47:53 AM11/9/09
to castalia-...@googlegroups.com
Dear All,
I was trying to evaluate the energy consumption for different mac in
Castalia using the valuePropagation example.
If I fixed the duty cycle and used different mac, TMAC, 14.5 with GTS
and basic CSMA . It seems that Tmac performs better in the energy
consumption . I thought that GTS should performs better. why is that?
have anyone done this before.?
Thank you in advance for any hints.
Kind regards
Sara

Athanassios Boulis

unread,
Nov 9, 2009, 6:03:00 PM11/9/09
to castalia-...@googlegroups.com

What do you mean "I fixed the duty cycle"? Some 
protocols, like T-MAC do not have/allow a fixed duty cycle.
The energy consumption of T-MAC is a direct consequence
of the adaptive/variable duty cycle, which in turn is
a function of the traffic you have.
You say that you expected GTS to be better. Do you know
what is the effective duty cycle with the parameters chosen?
If you left the parameters the same as the original distribution
then there is a 25% defined duty cycle and then because of 
GTS only 4 out of the 16 slots in the (potentially) active portion 
are getting used, so this is a 6.25% effective duty cycle.
This is possibly higher than what T-MAC can achieve under
low traffic conditions.
   
I believe you need to carefully study the MAC protocols 
(and their specific implementations if Castalia) you want to use.

Reply all
Reply to author
Forward
0 new messages