Ok, let's review just this much...I'm confused.
I think you have a system with three motes
one mote transmits messages at some rate -- Source
the second receives and re-transmits those messages -- Relay
the third is TOSBase, which receives (hopefully) everything -- Base
Your chart indicates that all Source messages are received at the Relay,
and that almost all messages, from everyone, are missed at the Base.
Is that what you meant to say? I assume that you have some tricky way
of determining how many messages are received at the Relay which doesn't
depend on them being received at the Base...
I think others have already said this but remember that only one
message can be on-the-air at any time, so the Source and Relay are
competing for air-time. If the Source is flooding the channel then
the two will compete for transmit windows. And, as I said in my
experience, the CC1000 backoff-retry isn't as good as one would hope
so messages go missing.
One experiment to try is to just measure the Source to Base message rate
with no Relay in the picture. And do that from both Source and Relay
positions to make sure you don't have some kind of multi-path interference
or other radio reception problem.
Also, I assume that the motes don't attempt to transmit a new message until
they have gotten a sendDone(). You shouldn't get the done() until the
backoff-retries have completed, and you should also be able to tell if the
transmit was successful or not.
MS
venkatesh s wrote:
Thanks for your kind reply Phil and Michael...
What about if i make the nodes to communicate in a single range, and
force them to come on hop., which means there shouldn't be hidden node
problems..
I did the following to achive the above, for the 3 node network..
Node 2 was forced to receive only node 1 packets (i.e, the route packets)
Node 1 was forced to receive both node 2 and node 0 (Rx) packets.
Node 0(Rx) was forced to receive only node 1 packets..
I used this logic in the event of Receive message in MultihopLEPSM.nc
file. so, this was routing the packets on hop but on a single range..
Under this case, i get the same results..
I thought of capturing why actually the packets are lost using a
TOSBase. It seems that TOSBase also missed some packets. I have
tabulated the results below, so that it will be clear..
ID | SentPackets | Packets received | Packet Drops | Packet Drops |
___|____________ | for forwarding | at Rx | at
TOSBase |
1 | 1000 | 996 | 1000
| 994 |
2 | 1000 | 0 |
994 | 997 |
__________________________________________________________ |
I really wonder over here, why there is no packet loss for node 1. I
experimented almost 10 - 12 times. It gave me almost the same results..
Few packets of node 2 were not received by node 1, while node 2 says he
has sent to medium.. Also, TOSBase was able to receive more packets of
node 2 than node1..
Is there such that the preamble getting corrupted which makes the packet
totally lost at node1??
------------------------------------------------------------------------
_______________________________________________
Tinyos-help mailing list
[email protected]
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
_______________________________________________
Tinyos-help mailing list
[email protected]
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
