Thanks you Tom and Marcus.
Yes, well... I meant that I've also checked FFT but a
picture of it was in another PC so I couldn't attach it when
I was writing the previous post.
Sorry about that. And I am currently using N210.
Today, carrying on from the yesterday, I've replaced a
square wave signal source block with either a random source
with a range of [0, 2) block or a custom source block from
my module
and made
In the figure above, I've marked a timing difference
between them with arrows.
Although I adjust a timing difference between them, a
waveform after the resampler seems ugly.
Yesterday, a waveform was quite acceptable since it's not
a (pseudo-)random signal, but an alternating square wave.
In this situation, a decision maker may make a false
decision, I think.
But it can be different if I connect USRP at the end of the
flow graph.
So I'll try both of them with various parameters, see the
results/differences and post something if I make some
progress.
Regards,