--- a/src/lte/doc/source/lte-testing.rst Fri Jun 08 11:12:00 2012 +0200
+++ b/src/lte/doc/source/lte-testing.rst Fri Jun 08 18:19:57 2012 +0200
@@ -368,7 +368,7 @@
--------------------
-The test suite ``lte-phy-error-model`` generates nine test cases with single eNB and a various number of UEs, all having the same Radio Bearer specification. Each test is designed for evaluating the error rate perceived by a specific TB size in order to verify that it corresponds to the expected values according to the BLER generated for CB size analog to the TB size. This means that, for instance, the test will check that the performance of a TB of :math:`N` bits is analogous to the one of a a CB size of :math:`N` bits by collecting the performance of a user which has been forced the generation of a such TB size according to the distance to eNB. In order to significantly test the BER at MAC level, we modified the Adaptive Modulation and Coding (AMC) module, the ``LteAmc`` class, for making it less robust to channel conditions by adding a configurable BER parameter (called ``Ber`` in the ns3 attribute system) which enable the selection of the desired BER at MAC level when choosing the MCS to be used. In detail, the AMC module has been forced to select the AMC considering a BER of 0.01 (instead of the standard value equal to 0.00005). We note that, these values do not reflect actual BER since they come from an analytical bound which do not consider all the transmission chain aspects; therefore the resulted BER might be different.
+The test suite ``lte-test-phy-error-model`` generates nine test cases with single eNB and a various number of UEs, all having the same Radio Bearer specification. Each test is designed for evaluating the error rate perceived by a specific TB size in order to verify that it corresponds to the expected values according to the BLER generated for CB size analog to the TB size. This means that, for instance, the test will check that the performance of a TB of :math:`N` bits is analogous to the one of a a CB size of :math:`N` bits by collecting the performance of a user which has been forced the generation of a such TB size according to the distance to eNB. In order to significantly test the BER at MAC level, we modified the Adaptive Modulation and Coding (AMC) module, the ``LteAmc`` class, for making it less robust to channel conditions by adding a configurable BER parameter (called ``Ber`` in the ns3 attribute system) which enable the selection of the desired BER at MAC level when choosing the MCS to be used. In detail, the AMC module has been forced to select the AMC considering a BER of 0.01 (instead of the standard value equal to 0.00005). We note that, these values do not reflect actual BER since they come from an analytical bound which do not consider all the transmission chain aspects; therefore the resulted BER might be different.
The parameters of the nine test cases are reported in the following:
@@ -403,7 +403,7 @@
BLER for test 6.
-The test verifies that in each case the expected number of packets received correct corresponds to a Bernoulli distribution with a confidence interval of 95%, where the probability of success in each trail is :math:`1-BER` and :math:`n` is the total number of packet sent.
+The test verifies that in each case the obtained number of packets received correctly falls within the 95% confidence interval of a Binomial distribution where the parameter :math:`p` is equal to :math:`1-BER`, the parameter :math:`n` is the total number of packets sent and the parameter :math:`k` is the number of packets correctly received.
MIMO Model