[Tinyos-help] Collection & gain

[Rémi Villé]

I made a test to just have one link between sink (which have no energy
problem) and the remaining of the node network, and data path always take de
path with less hop than possible, regardless to the gain...

So maybe this is not a problem...

I will make other tests, regardless to gain.

Rémi

2009/4/20 Rémi Villé <ville.remi at gmail.com>

> Hi,
>
> I used TOSSIM to see how reacts TinyOS Collection (CTP) according to the
> gain (thanks to MultihopOscilloscopeApp example).
>
> First, after some difficulties I understood that gain and noise are
> negative values (between 0 and -100), and more the value is near than -100
> more it's powerful (-100 for noise or 0 for gain prevent any communication).
>
> Second, I think that when gain becomes smaller than noise there's no
> possible communication.
>
> Am I wrong ?
>
> To have not any noise I choose -100 in every case.
>
> So I make some simulations to see the data path according to different gain
> values, with very good gain and very poor ones, but with a link between each
> nodes.
> And the data path don't change (the data go directly to the sink) !
>
> To see data paths I use "event bool  Intercept.forward(message_t* msg,
> void* payload, uint8_t len) " to add the intercepter node id in msg (and
> return TRUE in each case).
> In MultihopOscilloscopeAppC "MultihopOscilloscopeAppC.Intercept ->
> Collector.Intercept[AM_OSCILLOSCOPE];".
> The sink node print the path.
>
> So I wonder If Collection takes care of the gain, or only construct a tree
> with mote that can communicate with each other,
> so the routing gradient takes only care of paths with less hop than
> possible, and not faster ones ?
>
> I really don't understand, maybe I must search an energy reason... I will
> try other topologies...
>
> Thanks for help,
>
> Rémi
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.millennium.berkeley.edu/pipermail/tinyos-help/attachments/20090421/bfc9ab3b/attachment.htm