|
Hi:
I tried to trace rtt_ in Reno TCP to find how dose it change
when facing congestion in a bottle-neck link.
The result is:
0.40284 0 0 4 0 rtt_ 0.100
0.58770 0 0 4 0 rtt_ 0.200 1.09257 0 0 4 0 rtt_ 0.300 1.73257 0 0 4 0 rtt_ 0.700 5.16716 0 0 4 0 rtt_ 0.200 5.72716 0 0 4 0 rtt_ 0.400 6.76716 0 0 4 0 rtt_ 0.600 7.48716 0 0 4 0 rtt_ 0.700 8.36716 0 0 4 0 rtt_ 0.900 12.55689 0 0 4 0 rtt_ 0.200 13.38175 0 0 4 0 rtt_ 0.400 13.86175 0 0 4 0 rtt_ 0.500 14.50175 0 0 4 0 rtt_ 0.600 It seems that rtt_ is always a integer times of 0.1
.
To make the value of rtt_ preciser,I changed tcpTick in
ns-default.tcl from 0.1 to 0.001.
But the the value of rtt_ remained.
Who can tell me why,and how to make the value of rtt_ I traced
more accurate?
Thanks a lot!
wangyu |