Question regarding vector_sink_c

GNU Radio
1

Hi guys,

I’m using the vector_sink block to capture complex
data from the receiver USRP (in the hope to be able to
predict the wireless channel coeff) I printed the
recevied complex data using the rx_callback function
in receive_path.py script. The problem is the
vector_sink produces endless amount of complex number
even though my transmitter has already finish
transmitting. I connect my vector sink to
self.packect_receiver in receive_path.py, the vector
sink should be empty if no packet is received right?
If so why am i experiencing this problem. Please
advice. Thanks in advance.

  ____________________________________________________________________________________

Looking for last minute shopping deals?
Find them fast with Yahoo! Search.
http://tools.search.yahoo.com/newsearch/category.php?category=shopping

2

On Tue, Jan 08, 2008 at 08:26:43PM -0800, Wee S. wrote:

self.packect_receiver in receive_path.py, the vector
sink should be empty if no packet is received right?
If so why am i experiencing this problem. Please
advice. Thanks in advance.

It depends on where you’ve connected the vector_sink.

Note that vector_sink was really designed to be used only during QA
testing. It’s not generally useful because it ends up trying to store
a potentially infinite number of samples.

Eric

3

Hi eric,

Thanks for the reply, i have connected it to the
channel filters, i trying to do an estimate on
wireless channel coefficient. Therefore i’m trying to
capture the complex data received before
de-modulation. I havent have much success using vector
sink even when i set a range for the data captured,
all doesnt seems right.

Is there any blks that will be useful for my
experiment? Or do i have to write my own C++ blk to
process my needs. Thanks again.

regards,
ShinHan

— Eric B. [email protected] wrote:

in receive_path.py script. The problem is the

Eric

  ____________________________________________________________________________________

Be a better friend, newshound, and
know-it-all with Yahoo! Mobile. Try it now.
http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

4

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Wee S. wrote:

Is there any blks that will be useful for my
experiment? Or do i have to write my own C++ blk to
process my needs. Thanks again.

Look at blocks like gr.probe_*
(ingnuradio/gnuradio-core/src/lib/general/)

These will tell you some properties of the raw signal (namely,
amplitude). Something like that might be the kind of thing you want –
you could easily write a block in a similar style that computed some
other metric.

  • -Dan
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (GNU/Linux)
    Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHhXwmy9GYuuMoUJ4RAuInAJwO+boI3cmHIXW9S1c276Is3JcHCwCfdAYI
i8RL+rkXxQ6DZNnI/wQmKBE=
=uQ/3
-----END PGP SIGNATURE-----

5

Wee S. wrote:

Is there any blks that will be useful for my
experiment? Or do i have to write my own C++ blk to
process my needs. Thanks again.

regards,
ShinHan

The channel filter passes everything coming from the USRP and there’s no
squelch here, so you’ll get samples of the noise if you aren’t
transmitting. Even if you are transmitting from another USRP, the
received signal is going to be sampled at the wrong time and have
frequency and phase offsets. I’m guessing that’s what you mean by not
right.

Tom