F5/unison
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix documentation of the propagation loss test suite

Browse Source
This commit is contained in:
Tom Henderson
2009-10-04 22:56:35 -07:00
parent e3a6eefc88
commit 33026156cd
1 changed files with 19 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
doc/testing/propagation-loss.texi
View File

@@ -38,28 +38,24 @@ Distance :: Pr :: SNR
@subsection Validation test
Test program available online at: @uref{http://xxx.xxx.com,,}
Test program available in ns-3 at @code{src/devices/wifi/propagation-loss-model-test-suite.cc}
Taken at default settings (packetSize = 1000, numPackets = 1, lambda = 0.125, 802.11b at 2.4GHz):
Taken with values (lambda = 0.125m for 802.11b at 2.4GHz, SystemLoss = 1):
@smallformat
@verbatim
Distance :: Pr :: SNR
100 4.98265e-10W 1128.93
500 1.99306e-11W 45.1571
1000 4.98265e-12W 11.2893
2000 1.24566e-12W 2.82232
3000 5.53628e-13W 1.25436
4000 3.11416e-13W 0.70558
5000 1.99306e-13W 0.451571
6000 1.38407e-13W 0.313591
7000 1.01687e-13W 0.230393
8000 7.78539e-14W 0.176395
Distance :: Pr
100 4.98265e-10W
500 1.99306e-11W
1000 4.98265e-12W
2000 1.24566e-12W
@end verbatim
@end smallformat
@subsection Discussion
As can be seen, the SNR outputted from the simulator, and the SNR computed from the source's equation are identical.
As can be seen, the received power outputted from the model, and the power
computed from the source's equation are identical. The test suite tests
the expected value to the most significant digit of the input expected value.
@section LogDistancePropagationLossModel
@@ -98,31 +94,20 @@ Distance :: Pr :: SNR
@subsection Validation test
Test program available online at: @uref{http://xxx.xxx.com,,}
Test program available in ns-3 at @code{src/devices/wifi/propagation-loss-model-test-suite.cc}
Taken at default settings (packetSize = 1000, numPackets = 1, exponent = 3, reference loss = 46.6777, 802.11b at 2.4GHz)
Taken at default settings (exponent = 3, reference loss = 46.6777, 802.11b at 2.4GHz)
@smallformat
@verbatim
Distance :: Pr :: snr
10 4.98471e-9 11293.9
20 6.23089e-10 1411.74
40 7.78861e-11 176.468
60 2.30774e-11 52.2868
80 9.72576e-12 22.0585
100 4.98471e-12 11.2939
200 6.23089e-13 1.41174
500 3.98777e-14 0.0903516
1000 4.98471e-15 0.0112939
Distance :: Pr
10 4.98265e-9
20 6.22831e-10
40 7.78539e-11
80 9.73173e-12
@end verbatim
@end smallformat
@subsection Discussion
There is a ~.04% error between these results. I do not believe this is
due to rounding, as the results taken from the equation from the source
match exactly with the Friis results taken at one less power of ten.
(Friis and LogDistance can be modeled by Pt*Gt*Gr*lmb^2/((4*pi)^2*d^n*L),
where n is the exponent. n is 2 for Friis, and 3 for logDistance, which
accounts for the power of ten. ie: Friis at 100m is equivalent to LogDistance
at 10m.) Perhaps the ns-3 takes the random number into account despite
not being listed in the source.
As can be seen, the received power outputted from the model, and the power computed from the source's equation are identical. The test suite tests
the expected value to the most significant digit of the input expected value.