Re: [Discuss-gnuradio] [GSoC16] Signal intelligence proposal

[Sebastian Müller]

Sreeraj, thank you very much for your feedback! I updated my proposal on Github:

https://github.com/sbmueller/gsoc-proposal/blob/master/sigint-proposal.pdf

I have inserted a GUI draft for the frequency allocation information. Also, I added more detail

about the synchronization and when it is needed. I added a reference for modulation of unknown

signals and tried to make a new approach on the demodulation topic.

Since I won’t be home next week I can’t respond to any mails or update my proposal in this time.

But be ensured I will polish my proposal when I have more time in the week after, any further 

feedback is still welcome! Then I’m also planing to do some research on existing CNN 

possibilities in GR blocks.

Martin, I really like the idea of having two students working in this field. As I have pointed out in

my updated proposal, I have many ideas, maybe more than I can implement in 3 months. It

would be great if another mentor volunteered and gave a second student the possibility to 

work on this exciting topic.

Regards,

Sebastian

Am 11. März 2016 bei 12:02:26, sreeraj r (address@hidden) schrieb:

Hi Sebastian,

Good proposal. Really liked the radio service database idea too.
It would be nice to have a radio info tab in the GUI, which provide details about plausible transmission type based on the frequency allocation of the region (e.g [1]).

Some comments on your proposal

Block diagram (fig.2)

- In the block diagram your synchronization block comes before the AMC block. Are you referring to some blind synchronization schemes?. If yes can you put some references.

- You could also analyze whether sync is necessary for AMC.

- Some useful links [2-3].

Smart demodulation block

- Can you please add some more details about this? May be one or two references.

- For demodulation you need to do a lot more parameter estimation after AMC (samp/symbol, shaping filter parameters etc).

Modulation classification block
- It would be nice to add/analyze details on the existing gnuradio blocks for cyclo/CNN based classification

- This will help you to plan your timeline accordingly

[1] https://www.ntia.doc.gov/files/ntia/publications/2003-allochrt.pdf
[2] http://static1.1.sqspcdn.com/static/f/679473/25385817/1409511797677/rondeau-03-digital_demodulation.pdf?token=VLdlIIZq8xQmopwXjPf2Q5AgzkY%3D
[3] http://link.springer.com/chapter/10.1007%2F978-94-007-0107-6_20

Regards
Sreeraj Rajendran

On Thu, Mar 10, 2016 at 8:18 PM, Sebastian Müller <address@hidden> wrote:

Hi List!

My name is Sebastian Müller and I am highly interested in participating in this years GSoC program.

I want to take GSoC as the opportunity to begin with my participation in the GNU Radio project. Several of the ideas mentioned in the ideas list catched my attention, especially the topic on signal intelligence seems to be interesting for me, because I have some experience in this field. Also, improving the filter design tool would be a nice task for me.

I have taken some time the last few days to prepare my proposal for gr-sigint and evaluate approaches on this topic. Please find my proposal draft with more detailed information under

https://github.com/sbmueller/gsoc-proposal/blob/master/sigint-proposal.pdf

Feedback on my thoughts are very welcome!

Cheers,

Sebastian

_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio