[Smac-users] SMAC with adaptive listening saves much more energy than SMAC without adaptive listening ???

[Qasim Raza Iqbal]

Hi,
You could be right about a problem
I read that paper a few months ago. I know other protocols like T-MAC and
B-MAC, switch the radio off more quickly if no message is comming. Imagine
SMAC without adaptive listening, the radio will stay on a fixed period of
time and will then go in sleep mode depending on dutycycle. If a message
is present it will transmit otherwise it will stay idle. However, with
adaptive listening, if the radio comes on, and it has a short message
recieve timeout period, e.g that if it detects a message present , the
radio will stay on full time of the duty cycle mode, or otherwise it will
go sleep quicker than the dutycycle mode, and then come on again in the
next phase. Now I am not sure if that is present, but if it is, then it
could be saving more energy

what do you think

regards

Qssim Iqbal
UK


> Thanks, but the problem is that when you set the transition power to a low
> value, you will still find that SMAC with adaptive listening consumes much
> less energy than SMAC without adaptive listening. What's more, according
> to
> the SMAC paper, the energy consumption of both should approximately the
> same
> when the traffic is light. I think there are problems in SMAC code in
> NS2.29
> .
>    Thank you for your kind reply!
> Best regards
> David
>
>
> 2006/4/25, Qasim Raza Iqbal <iqbalqr at aston.ac.uk>:
>>
>> Hi
>>
>> The only obvious reason that I can see for that is, when the node goes
>> to
>> sleep, and then wakes up, The energy required for waking up (Transition)
>> is quite high. Imagine the node going to sleep 30 times a second and
>> waking up 30 times a second. The energy it will use waking up will
>> become
>> quite a lot over 3 or 6 months. However in adaptvie listening the node,
>> it
>> it can hear another packet coming, will stay awake and will recieve the
>> message and then go to sleep, so reducing the wakeup energy quite
>> considerably.
>>
>> what do you think?
>>
>>
>>
>>
>> > The values I got for the Transmission and receive power, the idle and
>> the
>> > sleep power  is from the data for the type of transceiver "WINS".
>> > What's more, I already set other values like the values in SMAC paper.
>> The
>> > result is that SMAC with adaptive listening consumes much less energy
>> than
>> > SMAC without adaptive listening even under light traffic. I use
>> NS2.29and
>> > I
>> > don't know why.
>> > Can you give me some suggestions? Thank you!
>> > Best regards!
>> > David
>> >
>> > 2006/4/24, Qasim Raza Iqbal <iqbalqr at aston.ac.uk>:
>> >>
>> >> Hi,
>> >>
>> >> I was wondering, were did you get the values for the Transmission and
>> >> receive power, THe idle and the sleep power. Are you trying to
>> implement
>> >> the powers that are used by a berkely Mote or some research paper. I
>> >> assume you got the timing values from the SMAC code itself.
>> >>
>> >> regards
>> >>
>> >> Qasim
>> >> ASton UK
>> >>
>> >>
>> >>
>> >>
>> >>
>> >> > Dear friends:
>> >> >     I use NS 2.29 and SMAC is included in it. I found that SMAC
>> with
>> >> > adaptive listening saves much more energy than SMAC without
>> adaptive
>> >> > listening even under light traffic. I am really confused about
>> this.
>> >> >    I have simulated SMAC with adaptive listening and without
>> adaptive
>> >> > listening.The simulation parameters are listed below:
>> >> > Radio bandwidth             20Kbps
>> >> > Channel coding              Manchester
>> >> > Control packet length          10bytes
>> >> > Data packet length(SYNC_CW)     up to 250 bytes
>> >> > MAC header length(DATA_CW)      10 byte
>> >> > Contention window for SYNC      31 slots
>> >> > Contention window for data     63 slots
>> >> > Slot Time                1ms
>> >> > DIFS                    10ms
>> >> > SIFS                    5ms
>> >> > Transmission power                    386 mw
>> >> > Receiving power                      368.2 mw
>> >> > Idle power                  344.2 mw
>> >> > Sleep power                            50 uw
>> >> > transition power                        50 mw
>> >> > transition time                       0.5 ms
>> >> > Duty cycle                10%
>> >> > Duration of listen interval           160ms
>> >> > the senario is a 11-node linear topology.The simulation lasts
>> 1000sec.
>> >> at
>> >> > time 50 sec, source node 0 began to send packets to sink node 10.
>> >> packet
>> >> > size is set to 50 bytes. There are totally 50 packets being sent.
>> >> >
>> >> > I calculated the total energy consumption of SMAC with adaptive
>> >> listening
>> >> > and without adaptive listening. When the cbr interval is 15 sec,
>> SMAC
>> >> > without adaptive listening consumed totally  1510.934236 J energy
>> >> while
>> >> > SMAC
>> >> > with adaptive listening consumed totally  863.737397 J
>> energy,nearly
>> >> only
>> >> > one half. I can't understand the result. When the cbr interval is
>> >> large,
>> >> I
>> >> > think the SMAC with and without adaptive listening should consume
>> >> > approximate the same energy. Do you think the result I obtained is
>> >> > reasonable?
>> >> >    any suggestions will be appreciated!
>> >> >    Thank you!
>> >> >                                                         your
>> sincere
>> >> >                                                            David
>> >> > _______________________________________________
>> >> > Smac-users mailing list
>> >> > Smac-users at mailman.isi.edu
>> >> > http://mailman.isi.edu/mailman/listinfo/smac-users
>> >> >
>> >>
>> >>
>> >>
>> >
>>
>>
>>
>