Joao/Delphine,  Clearly it depends on the data you're sending and the scenario 
you're sending it in.  If you need to send 29 bytes of information that were 
acquired in a single data acquisition, then 29 transmissions would cost you 
significantly more than one, 29 byte transmission for a number of reasons:1) 
bits of overhead per packet sent2) energy expended listening for time slot 
using csma 29 times vs 1 time (if your data is being continuously acquired, 
this depends on your architecture somewhat)3) energy to store bytes into 
non-volatile ROM (FLASH requires a higher voltage than normally operating 
microcontroller operations), Can your ucontroller sleep and keep it's volatile 
RAM - in tact? In general, I think the answer is "it depends".  Another issue 
is the data acquisition rate.  That is, if you're storing 32 points to wait to 
send it out vs your data is read 1 byte at a time and you send it out at one 
byte at a time.  Are there latency requirements, etc?Kurt--Forwarded Message 
Attachment--From: [email protected]: [email protected]: 
Wed, 17 Dec 2008 09:49:47 -0300Subject: Re: [Tinyos-help] Payload length 
influence on energy consumption    and padding. Delphine,          probably you 
will not see significant difference in energy consumption when you shorten the 
payload. As far as I am concerned, for default, the applications in TinyOS use 
BMAC protocol with a large preamble. Then, if you reduce payload length from 29 
bytes to 1 byte, maybe you will be able to see a reduction in energy 
consumption no more than 5%.                 David Culler, Philip Levis, and 
all TinyOS community pardon me if I am writing a wrong information !            
    In other words, could somebody confirm my informations ? *****************  
João Giacomin  Prof-DCC-UFLA*****************  ---------- Original Message 
-----------From: "Zainul M Charbiwala" <[email protected]>To: "delphine 
christin" <[email protected]>Cc: [email protected]: 
Thu, 11 Dec 2008 14:51:53 -0800Subject: Re: [Tinyos-help] Payload length 
influence on energy consumption and padding > Hi Delphine,> > Avrora tries to 
emulate the behavior of the chips as much as possible.> A Mica2 radio is byte 
based and I know of no routine in Avrora that> adds padding bytes. I believe 
the padding comes from TinyOS's packet> management. Try changing the value of 
TOSH_DATA_LENGTH in> tos/types/message.h to see the effect on padding.> > 
Zainul.> > On Thu, Dec 11, 2008 at 1:02 PM, delphine christin> 
<[email protected]> wrote:> > Hello,> >> >> > I would like to show that 
the payload length influences the radio energy> > consumption. For example, a 4 
byte payload requires less power to be sent> > than a 28 byte payload. In order 
to show it, I use the following scenario:> >> > -Boot.booted: get the data to 
send and switch the radio on> > -start.done:  send the data> > -send.done:  
stop the radio> > -stop.done: start a timer> > -timer.fired: start the radio> 
>> > I have already tested this scenario with Avrora and AEON. But 
unfortunately,> > the radio energy is constant for amount of data between 1 and 
10 bytes and> > then jumps to another value between 10 and 29 bytes.> > I have 
checked the packet with the -monitors=packets option and observe> > padding.> 
>> > (1) Is this padding specific to this simulator? or to TinyOS ? or both?> 
>> > (2) Has Power Tossim the same behavior concerning padding?> >> > (3) I 
foresee to use breakpoints and the energy given by the data sheets to> > obtain 
the consumption from the CPU and the radio. Do you have any advices> > or 
better ideas?> >> > Thanks a lot in advance,> >> > Best regards> >> > 
DelphineC> >> > ________________________________> > Votre correspondant a 
choisi Hotmail et profite d'un stockage quasiment> > illimité. Créez un compte 
Hotmail gratuitement !> > _______________________________________________> > 
Tinyos-help mailing list> > [email protected]> > 
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help> >> > 
_______________________________________________> Tinyos-help mailing list> 
[email protected]> 
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help------- 
End of Original Message -------  
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to