11 KiB
Audio Modem Communication Library
Description
This program can be used to transmit a specified file between 2 computers, using a simple audio cable (for better SNR and higher speeds) or a simple headset, allowing true air-gapped communication (via a speaker and a microphone).
The sender modulates an input binary data file into an 32kHz audio, which is played to the sound card.
The receiver side records the transmitted audio, which is demodulated concurrently into an output binary data file.
The process requires a single manual calibration step: the transmitter has to find maximal output volume for its sound card, which will not saturate the receiving microphone.
The modem is using OFDM over an audio cable with the following parameters:
- Sampling rate: 32 kHz
- Baud rate: 1 kHz
- Symbol modulation: BPSK, 4-PSK, 16-QAM ,64-QAM
- Carriers: 2-11 kHz
This way, modem may achieve 60kbps bitrate = 7.5 kB/s.
A simple CRC-32 checksum is used for data integrity verification on each 250 byte data frame.
Installation
Make sure that numpy and PortAudio v19 packages are installed (on Debian):
$ sudo apt-get install python-numpy portaudio19-dev
Clone and install latest version:
$ git clone https://github.com/romanz/amodem.git
$ pip install --user -e amodem
For graphs and visualization (optional), install matplotlib Python package.
For validation, run:
$ export BITRATE=48 # explicitly select high MODEM bit rate (assuming good SNR).
$ amodem-cli -h
usage: amodem-cli [-h] {send,recv} ...
Audio OFDM MODEM: 48.0 kb/s (64-QAM x 8 carriers) Fs=32.0 kHz
positional arguments:
{send,recv}
send modulate binary data into audio signal.
recv demodulate audio signal into binary data.
optional arguments:
-h, --help show this help message and exit
Calibration
Connect the audio cable between the sender and the receiver, and run the following scripts:
- On the sender's side:
~/sender $ export BITRATE=48 # explicitly select high MODEM bit rate (assuming good SNR).
~/sender $ amodem-cli send --calibrate
- On the receiver's side:
~/receiver $ export BITRATE=48 # explicitly select high MODEM bit rate (assuming good SNR).
~/receiver $ amodem-cli recv --calibrate
If BITRATE is not set, the MODEM will use 1 kbps settings (single frequency with BPSK modulation).
Change the sender computer's output audio level, until all frequencies are received well:
3000 Hz: good signal
4000 Hz: good signal
5000 Hz: good signal
6000 Hz: good signal
7000 Hz: good signal
8000 Hz: good signal
9000 Hz: good signal
10000 Hz: good signal
If the signal is "too weak", increase the sender's output audio level.
If the signal is "too strong", decrease the sender's output audio level.
If the signal is "too noisy", the SNR is probably too low: decrease the background noise or increase the signal (without causing saturation).
You can see a video of the calibration process here.
Usage
- Prepare the sender (generate a random binary data file to be sent):
~/sender $ dd if=/dev/urandom of=data.tx bs=10KB count=1 status=none
~/sender $ sha256sum data.tx
008df57d4f3ed6e7a25d25afd57d04fc73140e8df604685bd34fcab58f5ddc01 data.tx
- Start the receiver (will wait for the sender to start):
~/receiver $ amodem-cli recv -vv -i data.rx
- Start the sender (will modulate the data and start the transmission):
~/sender $ amodem-cli send -vv -o data.tx
- A similar log should be emitted by the sender:
2015-01-16 11:49:25,181 DEBUG Audio OFDM MODEM: 48.0 kb/s (64-QAM x 8 carriers) Fs=32.0 kHz amodem-cli:174
2015-01-16 11:49:27,028 INFO Sending 2.150 seconds of training audio send.py:63
2015-01-16 11:49:27,029 INFO Starting modulation send.py:68
2015-01-16 11:49:28,016 DEBUG Sent 6.000 kB send.py:50
2015-01-16 11:49:28,776 INFO Sent 10.000 kB @ 1.701 seconds send.py:73
- A similar log should be emitted by the receiver:
2015-01-16 11:49:24,369 DEBUG Audio OFDM MODEM: 48.0 kb/s (64-QAM x 8 carriers) Fs=32.0 kHz amodem-cli:174
2015-01-16 11:49:24,382 DEBUG Skipping 0.100 seconds recv.py:214
2015-01-16 11:49:24,535 INFO Waiting for carrier tone: 3.0 kHz recv.py:221
2015-01-16 11:49:26,741 INFO Carrier detected at ~1761.0 ms @ 3.0 kHz: coherence=99.944%, amplitude=0.499 detect.py:64
2015-01-16 11:49:26,741 DEBUG Buffered 1000 ms of audio detect.py:66
2015-01-16 11:49:26,912 DEBUG Carrier starts at 1761.000 ms detect.py:76
2015-01-16 11:49:26,917 DEBUG Carrier symbols amplitude : 0.499 detect.py:101
2015-01-16 11:49:26,917 DEBUG Current phase on carrier: -0.466 detect.py:112
2015-01-16 11:49:26,917 DEBUG Frequency error: -0.01 ppm detect.py:113
2015-01-16 11:49:26,917 DEBUG Frequency correction: 0.005 ppm recv.py:225
2015-01-16 11:49:26,917 DEBUG Gain correction: 2.004 recv.py:228
2015-01-16 11:49:27,099 DEBUG Prefix OK recv.py:48
2015-01-16 11:49:27,925 DEBUG 3.0 kHz: SNR = 40.58 dB recv.py:92
2015-01-16 11:49:27,925 DEBUG 4.0 kHz: SNR = 41.98 dB recv.py:92
2015-01-16 11:49:27,925 DEBUG 5.0 kHz: SNR = 42.81 dB recv.py:92
2015-01-16 11:49:27,925 DEBUG 6.0 kHz: SNR = 43.71 dB recv.py:92
2015-01-16 11:49:27,926 DEBUG 7.0 kHz: SNR = 43.43 dB recv.py:92
2015-01-16 11:49:27,926 DEBUG 8.0 kHz: SNR = 42.96 dB recv.py:92
2015-01-16 11:49:27,926 DEBUG 9.0 kHz: SNR = 42.66 dB recv.py:92
2015-01-16 11:49:27,926 DEBUG 10.0 kHz: SNR = 42.22 dB recv.py:92
2015-01-16 11:49:27,928 INFO Starting demodulation recv.py:119
2015-01-16 11:49:28,008 DEBUG Got 0.600 kB, realtime: 80.73%, drift: -0.00 ppm recv.py:140
2015-01-16 11:49:28,081 DEBUG Got 1.200 kB, realtime: 76.60%, drift: -0.00 ppm recv.py:140
2015-01-16 11:49:28,153 DEBUG Got 1.800 kB, realtime: 75.17%, drift: -0.00 ppm recv.py:140
2015-01-16 11:49:28,224 DEBUG Got 2.400 kB, realtime: 74.02%, drift: -0.00 ppm recv.py:140
2015-01-16 11:49:28,306 DEBUG Got 3.000 kB, realtime: 75.72%, drift: -0.00 ppm recv.py:140
2015-01-16 11:49:28,382 DEBUG Got 3.600 kB, realtime: 75.71%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,458 DEBUG Got 4.200 kB, realtime: 75.72%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,528 DEBUG Got 4.800 kB, realtime: 75.10%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,609 DEBUG Got 5.400 kB, realtime: 75.76%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,686 DEBUG Got 6.000 kB, realtime: 75.80%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,757 DEBUG Got 6.600 kB, realtime: 75.36%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,828 DEBUG Got 7.200 kB, realtime: 75.03%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,909 DEBUG Got 7.800 kB, realtime: 75.50%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:28,980 DEBUG Got 8.400 kB, realtime: 75.15%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:29,051 DEBUG Got 9.000 kB, realtime: 74.88%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:29,261 DEBUG Got 9.600 kB, realtime: 83.31%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:29,342 DEBUG Got 10.200 kB, realtime: 83.23%, drift: -0.01 ppm recv.py:140
2015-01-16 11:49:29,343 DEBUG EOF frame detected framing.py:57
2015-01-16 11:49:29,343 DEBUG Demodulated 10.205 kB @ 1.415 seconds (83.2% realtime) recv.py:165
2015-01-16 11:49:29,343 INFO Received 10.000 kB @ 1.415 seconds = 7.066 kB/s recv.py:169
- After the receiver has finished, verify the received file's hash:
~/receiver $ sha256sum data.rx
008df57d4f3ed6e7a25d25afd57d04fc73140e8df604685bd34fcab58f5ddc01 data.rx
You can see a video of the data transfer process here.
Visualization
Make sure that matplotlib package is installed, and run (at the receiver side):
~/receiver $ amodem-cli recv --plot -o data.rx
Donations
Want to donate? Feel free. Send to 1C1snTrkHAHM5XnnfuAtiTBaA11HBxjJyv.
Thanks :)