Post by Omar Rizwan (@omar@social.omar.website)

Omar Rizwan @omar

i want to post a video of current work but my server doesn't support video yet :')

Omar Rizwan @omar

Omar Rizwan @omar

Omar Rizwan @omar

github.com/analogdevicesinc/plutosdr-fw/issues/19

Omar Rizwan @omar

Omar Rizwan @omar

www.piclist.com/techref/uk/ac/york/cs/www-users/http/~fisher/mkfilter/index.htm

Omar Rizwan @omar

current status

Antennas and SDRs setup on cardboard

Antennas and SDRs setup on cardboard

Omar Rizwan @omar

Omar Rizwan @omar

Hmmmmmm

Omar Rizwan @omar

Whiteboard surface of table with a dot and "calibrate here (r = 915)" written in blue, with the side of a projected circle running through the dot

Whiteboard surface of table with a dot and "calibrate here (r = 915)" written in blue, with the side of a projected circle running through the dot

Omar Rizwan @omar

"pluto plus" search on AliExpress, first 4 results: first 3 are radio stuff, fourth is a Pluto from Disney doll

"pluto plus" search on AliExpress, first 4 results: first 3 are radio stuff, fourth is a Pluto from Disney doll

Omar Rizwan @omar

It somehow decays after a while and then starts working again after you sync 🤔

Omar Rizwan @omar

constellation

Omar Rizwan @omar

Omar Rizwan @omar

consistency coloring (the red flash is when i moved it) https://www.youtube.com/watch?v=eWROkMQ-6oc

Omar Rizwan @omar

typical desync

Omar Rizwan @omar

redoing sync/round stuff

Omar Rizwan @omar

right side of screen plastered with "Disk Not Ejected Properly" notifications for PlutoSDR

right side of screen plastered with "Disk Not Ejected Properly" notifications for PlutoSDR

Omar Rizwan @omar

Good morning!

as far as I can tell, basically all the problems in my setup are bc of the bigger amplifier. with the Nooelec, it seems very solid and probably on par with yours
but then obviously the range is short (and I also don't think we're getting enough multipath that the filtering or peak selection is really getting tested -- argmax + unfiltered phases work well on their own)
anyway, outstanding problems with 1D localization seem to be:

1. find a way to get back to 0.5W-1W output (a cleaner amp, or an output filter for the amp?) so we can get back to that ~10ft range

2. hidden Markov model or voting model or some other way to deal with +/-1 integer offset, which is showing up pretty consistently for me

3. some peak selection heuristic (especially once we get through #1 and start testing in more multipath situations)
overall, it seems like it's in decent shape. there are also the further directions of 2D (which I've started on -- it should mostly be plumbed in, but sync not working atm), higher throughput/OFDM, and RFID tag identification

Good morning!

as far as I can tell, basically all the problems in my setup are bc of the bigger amplifier. with the Nooelec, it seems very solid and probably on par with yours
but then obviously the range is short (and I also don't think we're getting enough multipath that the filtering or peak selection is really getting tested -- argmax + unfiltered phases work well on their own)
anyway, outstanding problems with 1D localization seem to be:

1. find a way to get back to 0.5W-1W output (a cleaner amp, or an output filter for the amp?) so we can get back to that ~10ft range

2. hidden Markov model or voting model or some other way to deal with +/-1 integer offset, which is showing up pretty consistently for me

3. some peak selection heuristic (especially once we get through #1 and start testing in more multipath situations)
overall, it seems like it's in decent shape. there are also the further directions of 2D (which I've started on -- it should mostly be plumbed in, but sync not working atm), higher throughput/OFDM, and RFID tag identification