search

naj1024 / pyspectrum

Python based spectrum analyser with web UI and sdr input
MIT License
24 stars 5 forks source link

readme

A Python/webUI RF spectrum analyser

Screenshot

This was an exercise in writing some python which expanded into providing a web based UI. The fft computations are done by libraries in Python, so not the fastest.

If you have some sort of sdr working with other tools then after installing the required python dependencies it should just work, mmmmm.

This project is a bit like the Krikkit One spaceship in The Hitchhiker's Guide to the Galaxy. It does the job, but not something you want to rely on.

Performance will depend on your machine and how the supporting fft libraries were compiled.

Features

Issues

TODO

Dependencies

The following python modules should be installed. Optional ones provide specific capabilities.

These should be installed into a virtual environment to not clash with system installed versions.

Required:
    numpy
    websockets
    matplotlib

Testing:
    pytest

Optional:
    scipy       - another FFT library
    pyfftw      - another FFT library, faster above 8k size (for me)
    pyadi-iio   - pluto device
    iio         - pluto device
    pyrtlsdr    - rtlsdr devices
    sounddevice - audio and funcube devices
    soapysdr    - soapy support
    paho-mqtt   - mqtt plugin (client)
    hid         - funcube control through usb hid, linux only?
    sigmf       - sigmf file support

Input

Data sources

Data types

Installation

Windows

Should be similar to the linux install, but

linux - also see pi install examples

cd ~
git clone https://github.com/naj1024/pyspectrum.git

Edit requirements.txt for required input sources.

vi ./pyspectrum/src/requirements.txt

I'm using pipenv for a virtual environment.

pipenv shell
pip install -r ./pyspectrum/src/requirements.txt

Run, then connect to localhost:8080 in a browser

python3 ./pyspectrum/src/pyspectrum.py

A complete install

Raspberry pi5

# From fresh install of bookworm on RP-5 (January 2024)
    $ sudo apt update
    $ sudo apt upgrade

# python virtual environement     
    $ sudo apt install pipenv
    $ pipenv shell
    (pi) pi@pi5:~ $ ctrl-d

    exited from shell - just needed virtual env configured

# rtlsdr hardware support, if required
    $ sudo apt install librtlsdr0
    $ sudo apt install librtlsdr-dev
    $ sudo apt install rtl-sdr
    $ sudo usermod -a -G plugdev pi

# pluto hardware support, if required
    $ sudo apt install libiio0 libiio-utils

# audio hardware support, if required
    $ sudo apt install libportaudio2

# hardware support for control support of funcube devices, if required
    $ sudo apt install libhidapi-hidraw0 libhidapi-libusb0

# sdrplay API support, required for any soapy builds. 
# Note API version can cause problems so get latest of everything
    ## Down load SDRplay_RSP_API-Linux-xxxxxxx.run from https://www.sdrplay.com/api/ 
    $ chmod +x SDRplay_RSP_API-Linux-3.14.1.run
    $ sudo ./SDRplay_RSP_API-Linux-3.14.1.run
    $ sudo systemctl status sdrplay
    $ sudo systemctl enable sdrplay

# Soapy because distro soapy does not include your device, some example builds given here
    ## dependancies for soapy build
    $ sudo apt install cmake g++ libpython3-dev python3-numpy swig

    ## sdrplay base, required
    $ cd ~
    $ mkdir soapy
    $ cd soapy
    $ git clone https://github.com/pothosware/SoapySDR.git
    $ cd SoapySDR
    $ mkdir build
    $ cd build
    $ cmake ..
    $ make
    $ sudo make install
    $ sudo ldconfig
    $ SoapySDRUtil --info    # no modules found

    ## Soapy sdrplay module, if required
    $ cd ~/soapy
    $ git clone https://github.com/SDRplay/SoapySDRPlay.git
    $ cd SoapySDRPlay
    $ mkdir build
    $ cd build
    $ cmake ..
    $ make
    $ sudo make install
    $ sudo ldconfig
    $ SoapySDRUtil --info    # libsdrPlaySupport.so module found

    ## test how fast raw driver can go, change sample rate 
    $ SoapySDRUtil --rate=8e6 --direction=RX --args="driver=sdrplay"

    ## Soapy rtlsdr module, if required
    $ cd ~/soapy
    $ git clone https://github.com/pothosware/SoapyRTLSDR.git
    $ cd SoapyRTLSDR
    $ mkdir build
    $ cd build
    $ cmake ..
    $ make
    $ sudo make install
    $ SoapySDRUtil --info    # librtlsdrSupport.so module found

    ## Soapy adalm pluto, if required
    $ sudo apt install libiio-dev
    $ cd ~/soapy
    $ git clone https://github.com/pothosware/SoapyPlutoSDR
    $ cd SoapyPlutoSDR
    $ mkdir build
    $ cd build
    $ cmake ..
    $ make
    $ sudo make install

    ## Soapy audio, if required - useful for testing if you have audio in
    $ sudo apt install librtaudio-dev
    $ cd ~/soapy
    $ git clone https://github.com/pothosware/SoapyAudio
    $ cd SoapyAudio
    $ mkdir build
    $ cd build
    $ cmake ..
    $ make
    $ sudo make install

# Soapy support
    ###########
    # Either allow system search paths for the python environment for soapy
    ######
    $ vi ~/.local/share/virtualenvs/pi-xxxxxx/pyvenv.cfg
        # allow search on system packages as Soapy has installed them there
        include-system-site-packages = true

    ###########
    # Or copy the relevant file into our local environment
    ######
    $ find /usr -name SoapySDR.py -ls
    $ find /usr -name _SoapySDR.so -ls
    # copy the files into your environment, pi-xxxxxx
    cp /usr/local/lib/python3.11/dist-packages/SoapySDR.py \
        /usr/local/lib/python3.11/dist-packages/_SoapySDR.so \ 
        ~/.local/share/virtualenvs/pi-xxxxxx//lib/python3.11/site-packages/

# let things sort themselves out, libraries, rules, groups etc
    $ sudo ldconfig
    $ sudo reboot

# test for working rtlsdr devices, as normal user
    $ rtl_test

# clone the pyspectrum repository
    $ cd
    $ git clone https://github.com/naj1024/pyspectrum.git
    $ cd pyspectrum
    $ git checkout flask

    ## install python requirements
    $ cd
    $ pipenv shell
    $ cd pyspectrum/src

    # edit for the features you require 
    $ vi requirements.txt
    $ pip install -r ./requirements.txt

    ## Run the spectrum anlayser
    $ python3 ./pyspectrum.py -vvv
    web server port 8080
    web socket port 8081
     * Serving Flask app 'webUI.FlaskInterface'
     * Debug mode: off

    ctr-c

    # check logs for errors
    $ cat ./logs/SpectrumAnalyser.log

#############
# Alternative soapy drivers 
#   - distrubution provided 
#   - no sdrplay
#######
$ sudo apt install soapysdr-module-all
$ sudo apt install python3-soapysdr 

$ vi ~/.local/share/virtualenvs/pi-xxxxxx/pyvenv.cfg
    # allow search on system packages as Soapy has installed them there
    include-system-site-packages = true

# check that path includes system path python3/dist-packages
$ pi@pi5:~ $ pipenv shell
Launching subshell in virtual environment...
pi@pi5:~ $  . /home/pi/.local/share/virtualenvs/pi-xxxxxx/bin/activate
(pi) pi@pi5:~ $ python3.11 -c "import sys; print('\n'.join(sys.path))"

/usr/lib/python311.zip
/usr/lib/python3.11
/usr/lib/python3.11/lib-dynload
/home/pi/.local/share/virtualenvs/pi-xxxxxx/lib/python3.11/site-packages
/usr/local/lib/python3.11/dist-packages
/usr/lib/python3/dist-packages
/usr/lib/python3.11/dist-packages
(pi) pi@pi5:~ $ exit

Soapy support on linux

The easiest way to get SoapySDR support is to install it from your distrbutions repository.

However, your distribution may not include the soapy driver you need, see above for source build instead.

$ apt install python3-soapysdr
$ dpkg -L dpkg -L python3-soapysdr

Then either let your virtual environment searches the system paths as well, say it is called sid-xxxx

$ vi ~/.local/share/virtualenvs/pi-xxxx/pyvenv.cfg
    # allow search on system packages as Soapy has installed them there
    include-system-site-packages = true

Or copy the files into the virtual environment

$ find /usr -name SoapySDR.py -ls
$ find /usr -name _SoapySDR.so -ls
$ cp /usr/local/lib/python3.11/dist-packages/SoapySDR.py \
        /usr/local/lib/python3.11/dist-packages/_SoapySDR.so \ 
        ~/.local/share/virtualenvs/pi-xxxxxx//lib/python3.11/site-packages/

Tested with

Windows: audio, file, pluto, rtlsdr, rtltcp, socket, funcube
Linux  : audio, file, pluto, rtlsdr, rtltcp, socket, funcube
         soapy(audio, rtlsdr, sdrplay)

On windows make sure you have the correct rlibrtlsdr.dll for your python 32bit/64bit

Command line Examples

FROM A PIEPENV SHELL PROMPT:

pipenv shell
cd pyspectrum/src
python ./pyspectrum.py         - Then goto http://127.0.0.1:8080 and configure the source

python ./pyspectrum.py -h      - help

python ./pyspectrum.py -i?     - list input sources that are available

python ./pyspectrum.py -vvv     - max debug output in logs/SpectrumAnalyser.log

Some default input selections, you normally select through web interface:

  python ./pyspectrum.py -ipluto:192.168.2.1 -c433.92e6 -s600e3   - pluto at 433MHz and 600ksps

  python ./pyspectrum.py -ipluto:192.168.2.1 -c433.92e6 -s1e6 
                          --plugin analysis:peak:threshold:12 
                          --plugin report:mqtt:broker:192.168.0.101     - detect and log signals

  python ./pyspectrum.py -ifile:test.wav -c433.92e6    - a test wav file

  python ./pyspectrum.py -iaudio:1 -s48e3 -iaudio:1    - audio input 

  python ./pyspectrum.py -irtlsdr:kk -c433.92e6 -s1e6   - rtlsdr

SOAPY:
  python ./src/pyspectrum.py -isoapy:audio -s48000 -c0  - soapy input
  python ./src/pyspectrum.py -isoapy:sdrplay -s2e6 c433.92e6

Debugging

Some useful tools for testing things under linux, e.g. transfer rates.

top     - process monitoring
iftop   - network bandwidth monitoring
usbtop  - USB bandwidth monitoring, needs usbmon kernel module installed
iostat  - device monitoring
dstat   - disk activity monitor
glances - monitoring tool
nmon    - monitoring tool

AD936x pluto XO support

We can add support for pluto frequency correction in ppm by adding the following to the file ad936x.py when pyadi-iio is installed. You should find the ad936x.py file under the adi directory in the site packages traversed by your environment. Insert the lines in the ad9364 class definitions.

@property
def xo_correction(self):
    return self._get_iio_dev_attr("xo_correction")

@xo_correction.setter
def xo_correction(self, value):
    self._set_iio_dev_attr_str("xo_correction", value)