See detailed introduction on my tech blog: Some efforts on GPS signal replay (GPS重放)
-------------*NEWS:----------------------------------------------------
ADS-B out is verified by HACKRF --> air --> rtl-sdr dongle. blog
See adsb directory and video:
http://youtu.be/xpLDqBkUiKc (out of China)
http://v.youku.com/v_show/id_XNzE4NzIzNDIw.html (in China)
---------------*Purpose:----------------------------------------------
Some efforts were put on replaying GPS signal, but it is only partially successful, not fully successful. Some tools/scripts, as side products, are useful. Learn many things from here: http://www.aholme.co.uk/GPS/Main.htm
WARNING：DO NOT SEND ANY REAL GPS SIGNAL. USE CHAMBER AND CLOSE-LOOP CABLE! MAKE SURE ALL YOUR OPERATIONS ARE LEGAL!
Get codes, program and sample GPS signals here: https://github.com/JiaoXianjun/GNSS-GPS-SDR
To build the C program "gps_test" in "c" directory (front part of software GPS receiver) after you get the whole repo:
cd c
make
To let this make smooth, make sure you have fftw library installed correctly (sudo apt-get install ...). There should be /usr/lib/libfftw.a.
a. generate GPS signal file by running gps_sig_gen.m in Matlab
gps_sig_gen.m
to get file gps_sig_tmp.bin
b. use "gps_test" program to do some initial demodulation and check C/A codes:
gps_test gps_sig_tmp.bin 2.046e6 8.184e6 5000
Above program gps_test performs initial GPS signal capture by multi-bins correlation in frequency domain.
gps_sig_tmp.bin: input GPS signal generated by gps_sig_gen.m (MATLAB).
2.046e6: carrier frequency
8.184e6: sampling rate
5000: maximum frequency offset for frequency searching
You will find that C/A codes results are aligned with GPS signal we generate by Matlab script.
1.2 use C program "gps_test" to receive GPS signal captured by others instead of my our Matlab script gps_sig_gen.m
gps_test gps.samples.1bit.I.fs5456.if4092.bin 4.092e6 5.456e6 5000
gps.samples.1bit.I.fs5456.if4092.bin: captured by others and can be downloaded here: http://www.jks.com/gps/gps.html
You will find C/A codes results are equal to those shown here: http://www.jks.com/gps/gps.html
2.1 playback gps.samples.1bit.I.fs5456.if4092.bin( http://www.jks.com/gps/gps.html )
a. run gps_bin1bit_log2bin.m in Matlab to generate playback file for HACKRF (they capture GPS signal with 1bit quantization. we convert it to 8bit format which is needed by HACKRF): gps.samples.8bit.IQ.fs5456.baseband.bin
b. run gps_Nottingham.grc to playback gps.samples.8bit.IQ.fs5456.baseband.bin over the air. And meanwhile/immediately,
c. use rtl-sdr to capture signal:
rtl_sdr -g 60 -f 1575.42e6 -s 2.8e6 -n 19.2e6 rtl_2.8Msps_1575.42MHz.bin
d. run proc_rtl_bin_for_gps('rtl_2.8Msps_1575.42MHz.bin') in Matlab to convert 8bit file to 1bit file which is needed by gps_test program：rtl_2.8Msps_1575.42MHz_1bit.bin
e. see C/A codes initial capture result by running gps_test program:
gps_test rtl_2.8Msps_1575.42MHz_1bit.bin 0.62e6 2.8e6 100000
You will find C/A codes results are equal to those shown here: http://www.jks.com/gps/gps.html
a. run gps.grc to playback gps_sig_tmp_for_hackrf_tx.bin over the air. And meanwhile/immediately,
b. capture air GPS signal by:
rtl_sdr -g 60 -f 1574.8e6 -s 2.8e6 -n 19.2e6 rtl_2.8Msps_1574.8MHz.bin
c. convert captured 8bit file to 1bit bin file which is needed by gps_test: rtl_2.8Msps_1574.8MHz_1bit.bin
proc_rtl_bin_for_gps('rtl_2.8Msps_1574.8MHz.bin'); (MATLAB)
d. RUN:
gps_test rtl_2.8Msps_1574.8MHz_1bit.bin 0.62e6 2.8e6 100000
to see GPS signal capture result: C/A codes are supposed to align with our expectation (Matlab script).
e. you can also capture air GPS signal with different intermediate frequency by:
rtl_sdr -g 60 -f 1575.42e6 -s 2.8e6 -n 19.2e6 rtl_2.8Msps_1575.42MHz.bin
f. convert captured 8bit file to 1bit bin file which is needed by gps_test: rtl_2.8Msps_1575.42MHz_1bit.bin
proc_rtl_bin_for_gps('rtl_2.8Msps_1575.42MHz.bin'); (MATLAB)
g. RUN:
gps_test rtl_2.8Msps_1575.42MHz_1bit.bin 0.62e6 2.8e6 100000
to see GPS signal capture result: C/A codes are supposed to align with our expectation (Matlab script).