My experiments with gnuradio.
- Open, and run
mrc.grc
. (looks like a play icon on the toolbar) This will compile then finish immediately. - Open, and run
phase_match.grc
. (looks like a play icon on the toolbar) This will compile then finish immediately. - Open, and run
fmaudiosink.grc
. (looks like a play icon on the toolbar) This will compile then finish immediately. - Use "Reload Blocks" (looks like two arrows in a circle icon on the toolbar)
- A new "E7EPL" section should appear on the right hand side, containing both blocks above, if not repeat the process above for the missing block.
- Open diversity2.grc and hit "Play".
- Remember to change the volume and source if you want to hear anything.
- It's likely you don't have the same SDR setup as me, you'll need to replace the SDR block with some other source.
See Wikipedia
I'm not sure I've quite got this right yet. I'm currently using the 2nd and 4th moments to estimate SNR. But it seems to work in the small test cases I've tried.
This takes two signals, cross correlates them, converts the correlation to a phase offset, and applies it to the signal so they should both be in phase.
This is for experiments around taking signals from two different antenna, phase matching them and combining them hopefully removing/reducing QRM (Interference), while increasing the signal.
Currently a work in progress.
This is just a trivial FM audio sink as one block, so I don't have to keep copying and pasting it everywhere.
This is used to create reproduciable test cases for experimentation.
The file written by this will be in cs16 format.
- c (complex. pairs of numbers representing the I/Q sample)
- s (signed) u (unsigned) or f (floating point)
- 8, 16, or 32 the bitwidth of the numbers in the pair.
They are just raw samples written to the file one after the other with no header, footer, framing, or checksum.
This is mostly used to make sure the data written out by the previous experiment worked and is reasonable.