discuss-gnuradio
[Top][All Lists]
Advanced

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

Re: [Discuss-gnuradio] reg overrun problem


From: Josh Blum
Subject: Re: [Discuss-gnuradio] reg overrun problem
Date: Mon, 30 Jan 2012 10:04:02 -0800
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:9.0) Gecko/20111229 Thunderbird/9.0


On 01/30/2012 09:41 AM, shantharam balasubramanian wrote:
> Hi
> 
> I was trying to run the benchmark_rx and benchmark_tx programs in the
> gnuradio 3.5.0 version, in my college lab. But I saw that when I am running
> the benchmark_rx in one of the wireless nodes, it shows the 'O'.
> 
> there are core i7 processors in the nodes, and I am not sure why this
> overrun problem is occurring. Can anyone tell me why this error occurs and
> what I should do to overcome it?
> 
> 

Ahh, le benchmark. Its more a proof of concept that you can implement a
mac layer in gnuradio, and less so a benchmark of anything. :-)

Here is a description of overflow:
http://files.ettus.com/uhd_docs/manual/html/general.html#overflow-underflow-notes

Faster processing can help, better code can help (if you care to
optimize benchmark rx*), and if you have a network product, larger
socket buffer can help:
http://files.ettus.com/uhd_docs/manual/html/transport.html#resize-socket-buffers

Not sure what modulation you selected, I think there may have been a
weird issue with the FLL. Try the GMSK modulation option (which doesnt
use the FLL), and see how well it works for you.

If you are feeling adventurous, I have SIMD-optimized float
adder/multiplier on my "next" branch. I think there are a few of these
multiplier blocks in the path of benchmark rx. So I am curious to hear
of the performance improvement of said math blocks provides any
noticeable effect for you in the overall benchmark rx app.
http://gnuradio.org/cgit/jblum.git/

Cheers,
-Josh



reply via email to

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