This is a couple of scripts to automatically find ZEVS transmismissions in a large dataset.
Please see: https://klubnl.pl/wpr/en/index.php/2022/06/16/analysis-of-zevs-messages/
The dataset is here: https://klubnl.pl/zevs/
It contains the following scripts:
parse_zevsdir.py - for each file in the dataset find which frequency that ZEVS transmits on has the strongest signal
find_zevs_transmissions.py - try to find valid ZEVS messages in the data generated by parse_zevsdir.py
plot.sh - plot the number of messages per day
spectrum.sh - using the files in the dataset, make a spectrogram for a given time range
generate_all_spectrograms.sh - generate a spectrogram for each message found by find_zevs_transmissions.py
runme.sh - run all of the above. Note: look inside this script to see how to download the ZEVS dataset
The scripts were intended to be ran on a modern linux system. Gnuplot and python3 must be installed.
To see how it works, read the runme.sh script. I hope to write better documentation someday.
This repository also has the results of running the scripts, so that you don't have to download the whole dataset and run it just to see the results:
zevs_freqs.txt - the result of running parse_zevsdir.py. contains the timestamp and frequency that ZEVS transmitted on
zevs_freqs_sort.txt - zevs_freqs.txt, sorted by timestamp
zevs_messages.txt - detected ZEVS messages
zevs_messages_perday.txt - number of ZEVS messages per day
zevs_messages_perday.pdf - a plot of messages per day, based on zevs_messages_perday.txt
zevs_activity.pdf - a plot of the number of minutes with 82Hz carrier active per day
spectrograms/ - a directory with a spectrogram of all messages in zevs_messages.txt (including false decodes).