M17 Modulator & Demodulator in C++ (GPL)
This program reads a 48k samples per second 16-bit, little-endian, single channel, M17 4-FSK baseband stream from STDIN and writes a demodulated/decoded 8k SPS 16-bit, single channel audio stream to STDOUT.
Some diagnostic information is written to STDERR while the demodulator is running.
This program reads in an 8k sample per second, 16-bit, 1 channel raw audio stream from STDIN and writes out an M17 4-FSK baseband stream at 48k SPS, 16-bit, 1 channel to STDOUT.
This code requires the codec2-devel, boost-devel and gtest-devel packages be installed.
This code also now requires the Blaze C++ math library.
It also requires a modern C++17 compiler (GCC 8 minimum).
mkdir build
cd build
cmake ..
make
make test
sudo make install
- Microsoft Visual Studio 2019
- Miniconda (or Anaconda) x64 for Windows
conda config --add channels conda-forge
conda create -n M17 vs2019_win-64 cmake ninja pkg-config boost-cpp gtest gmock gtest libcodec2
conda activate M17
This program was designed to be used with RTL-SDR, specifically rtl-fm.
rtl_fm -E offset -f 144.91M -s 48k | m17-demod -l | play -b 16 -r 8000 -c1 -t s16 -
You should run this in a terminal window that is 132 characters wide. It will output diagnostic information on a single line in the window.
sox ~/m17-demodulator/brain.wav -t raw - | ./m17-mod -S WX9O | ./m17-demod -l -d | play -q -b 16 -r 8000 -c1 -t s16 -
The input audio stream must be 1 channel, 16-bit, 8000 samples per second.
Use -S <callsign>
to set your source (callsign).
Use -b
to output a bitstream rather than baseband.
Use -h
to see the full help. Many of the options do not yet work.
The output of the modulator is 48ksps, 16-bit, 1 channel raw audio.
To output a bitstream file:
sox ~/m17-demodulator/brain.wav -t raw - | ./m17-mod -S WX9O -b > m17.bin
This bitstream file can be fed into m17-gnuradio to transmit M17 using a PlutoSDR (or any SDR with an appropriate GNU Radio sink), or loaded into a vector signal generator such as an ESG-D Series signal generator.
I have recently been testing the modulator and demodulator with long-form programs. One that works well is the AR Newsline program. This is available as a downloadable MP3 file. To convert that to a bitstream, run the following:
ffmpeg -i ~/Downloads/Report2287.mp3 -ar 8000 -ac 1 -f s16le -acodec pcm_s16le - | ./apps/m17-mod -b -S NS9RC > Report2287.bin
This can then be used with the the M17_Impaired.grc
GNU Radio flow graph.
There are two command line options for the demodulator that determine the diagnostic output.
-d causes demodulator diagnosts to be streamed to the terminal on STDERR.
-l causes the link setup frame information (either from the first frame or LICH) to be displayed.
Note that the oscillators on the PlutoSDR and on most RTL-SDR dongles are rather inaccurate. You will need to have both tuned to the same frequency, correcting for clock inaccuracies on one or both devices.
Also note that I need to use -E offset
to decode the data well, even though
I see this in the rtl_fm
output:
WARNING: Failed to set offset tuning.
This was tested using the m17-gnuradio GNU Radio block feeding an Analog Devices ADALM Pluto SDR, modulating the m17.bin file from the m17-demodulator repo.
The demodulator diagnostics are calibrated for the following command:
rtl_fm -F 9 -f 144.91M -s 18k | sox -t s16 -r 18k -c1 - -t raw - gain 9 rate -v -s 48k | ./apps/m17-demod -d -l | play -b 16 -r 8000 -c1 -t s16 -
Specifically, the initial rate (18k samples per second) and the conversion rate and gain (gain of 9, output rate 48k samples per second) are important for deviation and frequency offset.
The demodulator produces diagnost output which looks like:
SRC: BROADCAST, DEST: MBLKDTNC3, TYPE: 0002, NONCE: 0000000000000000000000000000, CRC: bb9b
dcd: 1, evm: 13.27%, deviation: 0.9857, freq offset: 0.03534, locked: true, clock: 1, sample: 0, 0, 0, cost: 9
The first line shows the received link information. The second line contains the following diagnostics.
- DCD -- data carrier detect -- uses a DFT to detect baseband energy. Very good at detecting whether data may be there.
- EVM -- error vector magnitude -- measure the Std Dev of the offset from ideal for each symbol.
- Deviation -- normalized to 1.0 to mean 2400Hz deviation for the input from the following command:
rtl_fm -F 9 -f 144.91M -s 18k | sox -t s16 -r 18k -c1 - -t raw - gain 9 rate -v -s 48k
-- the rates and gain are the important part. - Frequency Offset -- the estimated frequency offset from 0 based on the DC level of each symbol. The magnutude has not been measure but should be around 1.0 == 2kHz using the same normalized input as for deviation. Anything > 0.1 or less than -0.1 requires review/calibration of the TX and RX frequencies.
- Locked -- sync word detected.
- Clock -- estimated difference between TX and RX clock. Will not work if > 500ppm. Normalized to 1.0 -- meaning the clocks are equal.
- Sample -- estimated sample point based on 2 clock recovery methods. Should agree to within 1 always. There are 10 samples per symbol. Should never jump by more than 1 per frame. The third number is the "winner" based on certain heuristics.
- Cost -- the normalized Viterbi cost estimate for decoding the frame. < 5 great, < 15 good, < 30 OK, < 50 bad, > 80 you're hosed.
When transmitting a BER test, the diagnostics line will show additional information.
dcd: 1, evm: 2.626%, deviation: 0.9318, freq offset: -0.154, locked: true, clock: 1, sample: 2, 2, 2, cost: 2, BER: 0.000000 (53190 bits)
The BER rate and number of bits received are also displayed.
Thanks to the M17 team to for the great work on the spec.