Hi,

I've been playing with the TDMA model and wanted to gauge the results of moving 
nodes around in the demo.  The demo is 10 nodes all originally at approximate 
40 degrees latitude +/-and -74.5 degrees longitude +/- with altitude fixed at 
3.00.  I started out by moving just a single node, node 1, and fixed it at that 
location for the duration of the demo.  I noted when the node might disconnect, 
if at all.  Since this is a precomputed pathloss model with a specified value 
of 90 for all nodes, I expected the OLSRLINKVIEW to remain mutually connected 
through the duration, no matter the location of node 1.  However, it looks as 
though node 1 loses connectivity to all other nodes when I move it at or beyond 
-79.4 degrees longitude.  This, according to my calculations, looks to 
approximately correspond to the distance/time of flight for the specified slot 
duration of 1500 usec.  -79.3 remains connected and -79.5 certainly 
disconnects.  (I didn't change the values of latitude and longitude for 
remaining 9 nodes that were originally specified but they move only slightly 
from their original positions, still roughly approximating the 1500 usec 
distance with the 1500 usec slotduration.)  I revised the slotduration to 1600 
usec on a separate run and observed that the model completely reconnected.  I 
suspect that if I moved node 1 longitudinally an additional amount to 
correspond to 1600 usec time of flight, it too would disconnect in a like 
manner and indeed that is the case.

Is there a particular reason that the model would lose connectivity when 
distances exceed the specified slotduration?  Since the pathloss is fixed 
value, I was expecting to be able to extend well beyond that and still retain 
connectivity, although impractical in the real world, but trying to achieve a 
related objective.  The TDMA model staggers the tx slots from 1 to 10 and they 
don't appear to present any collision opportunity.  Have I overlooked something?
_______________________________________________
emane-users mailing list
[email protected]
https://publists.nrl.navy.mil/mailman/listinfo/emane-users

Reply via email to