Two Ardusimple Multiband GNSS antennas are fitted to the MAX2771-based PocketSDR clone. One antenna is facing the sky at an angle of about 45 degrees in the direction NISAR is beaming from (East or West depending on ascending or descending pass) while the second antenna is vertically mounted and facing in the opposite direction. The ground plane of the reference antenna shields the surveillance antenna from the direct signal.
Configuring the PocketSDR to record a 24 MHz bandwidth around 1229 MHz:
sudo PocketSDR/app/pocket_conf/pocket_conf ./pocket_NISAR_24MHz.conf
sudo PocketSDR/app/pocket_conf/pocket_conf
# [CH1] F_LO = 1229.025000 MHz, F_ADC = 24.000000 MHz (IQ), F_FILT = 0.0 MHz, BW_FILT = 23.4 MHz
# [CH2] F_LO = 1229.025000 MHz, F_ADC = 0.000000 MHz (IQ), F_FILT = 0.0 MHz, BW_FILT = 23.4 MHz
sudo PocketSDR/app/pocket_dump/pocket_dump /tmp/1.bin /tmp/2.bin
The recording only started a few seconds before the maximum elevation time predicted by Heavens Above for NISAR at our location (47.25N, 6E)
$ stat /tmp/1.bin
File: /tmp/1.bin
Size: 2142240768 Blocks: 4184064 IO Block: 4096 regular file
Device: 0,39 Inode: 180679 Links: 1
Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2026-03-26 20:03:05.885563149 +0100
Modify: 2026-03-26 19:51:43.824067742 +0100
Change: 2026-03-26 19:51:43.824067742 +0100
Birth: 2026-03-26 19:50:59.190699895 +0100
$ stat /tmp/2.bin
File: /tmp/2.bin
Size: 2142240768 Blocks: 4184064 IO Block: 4096 regular file
Device: 0,39 Inode: 180680 Links: 1
Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2026-03-26 19:50:59.190699895 +0100
Modify: 2026-03-26 19:51:43.824067742 +0100
Change: 2026-03-26 19:51:43.824067742 +0100
Birth: 2026-03-26 19:50:59.190699895 +0100
but still allowed to collect useful data on the surveillance and reference channels.
Because the MAX2771 provides the complex conjugate of the IQ stream (i.e. I-jQ) the original image was upside down when interpreting the file record content as I+jQ
and flipped after correcting for I-jQ as done in the nisar_process5.m:
Projecting on the map+DEM validates the source of the reflections
Since the MAX2771 samples with low resolution ADCs, attempts were made to replace the reference channel with synthetic chirps, requiring to match the time delay and phase of the observed signal: see the nisar_process6.m script aimed at performing such a computation
