Le 25 juin 2009 18:17, Philip Levis <[email protected]> a écrit :

>
> On Jun 18, 2009, at 5:05 AM, Wojciech Bober wrote:
>
>  Hi Remi,
>>
>> Maybe you can work something out from this slides:
>> http://www.stanford.edu/class/ee359/lectures.html
>>
>> Regards,
>> -- Wojtek
>>
>
> Please note all of these are formalized approximations/models for the
> purpose of analysis. Path loss is of course not a simple equation, due to
> things like multipath, humidity, temperature, node orientation, etc.
> Otherwise, ranging with raw RSSI would be trivial, contrary to the decade of
> research on the topic.
>
> Phil
>

Yes, the reading of these documentations make me understand how the radio
communication behaviour can be complicated to predict in real world. For
these reasons and other ones, I'm convince that real experimentation is a
crucial step and I don't say the contrary. But for my generator of mote
topology (for TOSSIM) I just would like to have a simplified function to
compute gain from distance.
For now, I don't consider multipath, obstacle, orientation, etc... I only
assume to dispatch motes in a large room.

So I say that if a mote is very close to an other one the gain is 0
(assuming that tx power is 0dB).
If the mote is in the radio range I use : Gain = -10*n*log(dist(A,B)) (with
n=4)
if the mote is not in the radio range then I just don't create a link
between them.

So it's very rough, but it's better than just choosing a randomly gain (I
prefer adding some random with the noise file of TOSSIM) without considering
distance, maybe should I have a function more proportional to the radio
range, or taking into account other basic parameters I forget... any help is
welcome.
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to