Hello list, I am using Octave to process time-series samples from my USRP N210. When I compute the periodogram of these time samples, I always get strange peaks at the edges of my PSD and I can definetly count out that they belong to the actual spectrum. Does anybody know what causes those peaks? (I dont think that this problem is caused by Octave, and this is why I am posting this question to the usrp-list and not to tho octave-list.) Regards Sebastian

On 12/05/2011 04:50 AM, Sebastian Döring wrote: > Hello list, > > I am using Octave to process time-series samples from my USRP N210. > > When I compute the periodogram of these time samples, I always get > strange peaks at the edges of my PSD and I can definetly count out that > they belong to the actual spectrum. > Does anybody know what causes those peaks? > > (I dont think that this problem is caused by Octave, and this is why I > am posting this question to the usrp-list and not to tho octave-list.) > At the start of burst, there will be 20 samples or so of transients due to the filters in the decimation chain. Are you perhaps seeing that? -josh

On Mon, 05 Dec 2011 07:41:15 -0800 Josh Blum <josh@ettus.com> wrote: > > > On 12/05/2011 04:50 AM, Sebastian Döring wrote: >> Hello list, >> >> I am using Octave to process time-series samples from my >>USRP N210. >> >> When I compute the periodogram of these time samples, I >>always get >> strange peaks at the edges of my PSD and I can definetly >>count out that >> they belong to the actual spectrum. >> Does anybody know what causes those peaks? >> >> (I dont think that this problem is caused by Octave, and >>this is why I >> am posting this question to the usrp-list and not to tho >>octave-list.) >> > > At the start of burst, there will be 20 samples or so of >transients due > to the filters in the decimation chain. Are you perhaps >seeing that? > > -josh > This could be an explanation, although there are rather less than 10 samples that form the peak and there´s also a smaller peak at the end of the spectrum. Thanks Sebastian > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Sebastian Döring <sdoering@...> writes: > > Hello list, > > I am using Octave to process time-series samples from my > USRP N210. > > When I compute the periodogram of these time samples, I > always get strange peaks at the edges of my PSD and I can > definetly count out that they belong to the actual > spectrum. > Does anybody know what causes those peaks? > > (I dont think that this problem is caused by Octave, and > this is why I am posting this question to the usrp-list > and not to tho octave-list.) > > Regards > Sebastian > Hello list, I am very sorry I may not have answers to Josh Blum's questions, because I am a complete stranger in that regard. I want to find out if Octave is capable of doing the same function that Simulink does with USRP 2. I means We can interface Simulink with UHD, so that we can communicate directly with the USRP hardware, instead of using the GNU radio software. Is Octave also capable of doing that? Do simulink blocks exist in Octave as well? Regards, Ken Kenn

Kenn - No, Octave is an open-source alternative to Matlab, but not to Simulink. http://www.gnu.org/software/octave/ Cheers, Ben ---------------------------- Ettus Research, LLC <http://www.ettus.com/> | USRP <http://goo.gl/Ixjhh> On Wed, Jun 12, 2013 at 8:03 AM, Kenn Kenn <alexcaundy@gmail.com> wrote: > Sebastian Döring <sdoering@...> writes: > > > > > Hello list, > > > > I am using Octave to process time-series samples from my > > USRP N210. > > > > When I compute the periodogram of these time samples, I > > always get strange peaks at the edges of my PSD and I can > > definetly count out that they belong to the actual > > spectrum. > > Does anybody know what causes those peaks? > > > > (I dont think that this problem is caused by Octave, and > > this is why I am posting this question to the usrp-list > > and not to tho octave-list.) > > > > Regards > > Sebastian > > > > > Hello list, > > I am very sorry I may not have answers to Josh Blum's questions, because I > am a complete stranger in that regard. I want to find out if Octave is > capable of doing the same function that Simulink does with USRP 2. I means > We can interface Simulink with UHD, so that we can communicate directly > with > the USRP hardware, instead of using the GNU radio software. Is Octave also > capable of doing that? Do simulink blocks exist in Octave as well? > > Regards, > Ken Kenn > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >