David Protzman
3b30b48402
Removed unused code from create_zc (MATLAB)
2022-10-09 18:00:25 -04:00
David Protzman
19ab4e3e35
Updated hash for utils class
2022-10-09 18:00:15 -04:00
David Protzman
b28013d211
Updated tests for utils to use a single instance of MATLAB
...
There were issues before where the single MATLAB instance would cause a segfault at the end of the test
Now it truly uses a single instance
2022-10-09 17:59:05 -04:00
David Protzman
4a493badb1
Removed constructor and deconstructor as these are set to default now
2022-10-09 17:58:43 -04:00
David Protzman
c3c8320c8f
Added comments for utils class
2022-10-09 17:58:10 -04:00
David Protzman
4ab63eee2e
Added GNU Radio module
2022-10-09 11:27:20 -04:00
David Protzman
ea67d38d43
Added ignore file
2022-10-09 11:27:15 -04:00
David Protzman
96e93aa3da
Updated create_zc to use zadoff_chu
...
This was done to make testing easier as there were bugs in the C++ that were painful to figure out since the raw ZC sequence was generated in this function
2022-10-09 11:22:24 -04:00
David Protzman
b700e24480
Added function for generating raw ZC sequences
2022-10-09 11:22:06 -04:00
proto17
5f1a468159
Create README.md
2022-09-22 22:56:45 -04:00
proto17
fd96323c50
Create LICENSE
2022-09-06 08:47:02 -04:00
proto17
4d83bd6267
Update README.md
...
Fixed list of frequencies that DroneID appears on
2022-09-05 16:56:00 -04:00
proto17
9e2bd4fe55
Removed unnecessary seek into the ZC sym for CFO
...
It was pointed out that the additional 6 sample seek wasn't necessary. Tested with a HackRF and DJI Mini 2 to validate.
2022-09-05 12:23:18 -04:00
David Protzman
a58848eaff
Merge branch 'main' of https://github.com/proto17/dji_droneid into main
2022-08-28 09:36:00 -04:00
David Protzman
5f319c4cc4
Moved files around to prevent clutter
...
Files that are not currently being used have been moved to unused_scripts
2022-08-28 09:34:45 -04:00
David Protzman
deb0be39c2
Removed normalization requirement
...
There's no reason it must be normalized, and Octave doesn't support normalization with xcorr
2022-08-28 09:21:18 -04:00
David Protzman
363ff70239
Small change in how the symbols are extracted
2022-08-28 01:17:33 -04:00
David Protzman
3c77c3e6a8
Added function that finds the start time offset based on the cyclic prefix
2022-08-28 01:16:23 -04:00
David Protzman
a731a12660
*Lots* of changes
...
These changes have been over several months and can't really be split apart meaningfully :\
2022-08-28 01:10:24 -04:00
David Protzman
fe761e9473
The equalizer takes care of global phase offsets
2022-08-28 01:08:25 -04:00
David Protzman
95bc95e236
Added integer frequency offset correction logic
2022-08-28 01:07:46 -04:00
David Protzman
339f1fcae9
Setting title of figure 43
2022-08-28 01:07:24 -04:00
David Protzman
ef608611d6
Now plotting the power spectrum with averaging and correct freq range
2022-08-28 01:06:39 -04:00
David Protzman
779e170785
Fixed issue with plots always showing up in burst extractor
2022-08-28 01:04:25 -04:00
David Protzman
9e313dd7b4
Removed example file values, moved some parameters
2022-08-28 01:00:42 -04:00
proto17
7fa0fc4988
Create est_integer_freq_offset.m
2022-05-18 23:40:07 -04:00
proto17
a170391a9e
Create test_integer_freq_offset.m
2022-05-18 23:30:40 -04:00
proto17
ec1ca41c6b
Create cyclic_prefix_sto_test.m
2022-05-17 22:37:01 -04:00
proto17
95863425aa
Create ifo_test.m
2022-05-16 23:19:27 -04:00
proto17
ca5fada2d0
Update README.md
2022-04-30 12:27:46 -04:00
proto17
f312a1cc14
Update README.md
2022-04-30 12:26:42 -04:00
proto17
b80be8d9b8
Update README.md
2022-04-30 12:25:51 -04:00
David Protzman
9e2aafab50
Moved the offset index back 1 FFT size
...
This is because the new xcorr function is forward looking vs reverse looking
2022-04-30 12:00:04 -04:00
David Protzman
46b2c4dad5
Warning the user if correlation threshold > 1.0
...
Now that the xcorr is truly normalized, there's no good reason to go above 1.0
2022-04-30 11:59:32 -04:00
David Protzman
388d4acd7b
Erroring if correlation threshold < 0
2022-04-30 11:59:05 -04:00
David Protzman
4309068874
Added comments
2022-04-30 11:58:21 -04:00
David Protzman
60ea5e6f3f
Shrunk the peak search window from 100 to 10
...
No reason to look that far around for the peak of a ZC sequence correlation
2022-04-30 11:58:02 -04:00
David Protzman
b442f73baf
Now using a truly normalized cross correlation
2022-04-30 11:57:09 -04:00
David Protzman
e4a49306b2
Added running sum for abs squared
2022-04-28 23:04:50 -04:00
David Protzman
9daf1ecaab
Added missing update to `prev_val`
2022-04-28 22:46:08 -04:00
David Protzman
1d2698c0af
Added a new normalized xcorr function that's faster
...
Does trade some accuracy in the process
2022-04-28 22:43:57 -04:00
David Protzman
63e26007f8
Merge branch 'main' of github.com:proto17/dji_droneid into main
2022-04-26 22:25:44 -04:00
David Protzman
c173de56b4
Added default sample type
2022-04-26 22:18:38 -04:00
David Protzman
5e800d24d0
Making use of get_bytes_per_sample function in convert file type
2022-04-26 22:17:53 -04:00
David Protzman
38352ec47b
Casting bursts to double complex
2022-04-26 22:17:17 -04:00
David Protzman
43a19fe44a
Added comments
2022-04-26 22:16:08 -04:00
David Protzman
863bf910be
Added varargs for extract bursts and find zc seq funcs
2022-04-26 22:15:47 -04:00
David Protzman
94ee1b3602
Updated get_sample_count_of_file to use new function for num of bytes/sample
2022-04-26 21:28:59 -04:00
David Protzman
3213764d68
Added function to get the number of bytes in a numeric type
2022-04-26 21:23:16 -04:00
David Protzman
df18420eaf
Fixed bug where floating point samples came back as 'single'
...
This would cause lots of functions to fail since a lot of the dsp functions rely on doubles
2022-04-26 21:12:32 -04:00