discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Save all the samples of a packet (all complex val


From: Jason Uher
Subject: Re: [Discuss-gnuradio] Save all the samples of a packet (all complex values) and demodulate at once. (Is it possible?)
Date: Wed, 2 Mar 2011 08:02:49 -0600

On Wed, Mar 2, 2011 at 6:00 AM, Minsuk Kang <address@hidden> wrote:
> Dear all,
> We're using benchmark_tx(rx) apps to implement a simple cellular-like
> network.
> In the benchmark_rx example, the demodulator demodulates some amount of
> symbols at a time and the correlator and the framer, which followed the
> demodulator, read the demodulated bits and packetize them into a frame.
> But, for some reasons, we want to change the way the receiver decodes a
> packet.
> That is, we are trying save all the samples of a packet (all complex
> values) and demodulate at once in a function at a receiver.
> However, we haven't implemented or seen this kind of receiver structure in
> gnuradio.
> If you have seen or any experience with similar examples, please let us
> know.
> Any comment on it would be very appreciated.
> Thanks in advance.
> PS. we're using USRP1 / gnuradio 3.3.0 / Ubuntu 10.04

The main issue with this is "what is a packet".  In the benchmark
scripts there is a defined start and end of the packet, namely,
correlation sequence at the beginning, and the n_bytes field.  To save
the demodulated samples into packets would not do you any good, as you
have no idea where packets begin and end.  You could just save fixed
size blocks using the packetizer, but then your real packets would be
split between your saved blocks at random.

If what you are trying to do is really the best way to go about it,
you would still need to demodulate at least /some/ of the bits
(probably just the header) so you would know when to a packet starts,
then packetize the raw stream starting at that point.

Jason



reply via email to

[Prev in Thread] Current Thread [Next in Thread]