This GNU Radio module is developed to realize signal analysis abilities in typical block-structure. The module is capable of the following:
- Energy detection of continuous signals
- Filtering of detected signals
- OFDM parameter estimation (carrier spacing, symbol time)
- Blind OFDM synchronization
- Resampling of signals to constant rate
- 3D Visualisation of FAM data, from gr-specest
- Using TensorFlow models for AMC
To install the Inspector, the following dependencies are required:
- GNU Radio
- gr-analog, gr-blocks, gr-channel, gr-filter
- QT4
- QWT 6.1.0
- QwtPlot3D
- TensorFlow 0.12 (pip installation - you need the version for Python 2.7)
Install by the following shell commands:
git clone git@github.com:gnuradio/gr-inspector.git
cd gr-inspector
mkdir build
cd build
cmake ..
make -j4
sudo make install
Get inspired by the examples. There are flowgraphs for each block that should explain how to use them.
For detailed information about each block, visit the Doxygen HTML pages.
A typical inspector flowgraph could look like this:
Performs an energy detection of continuous input signals. A threshold in dB can be set or an autmatic threshold calculation can be used.
GUI where the spectrum is displayed along with signal markers for the detected signals. Also, signal can be selected manually in this block.
Takes the signal info from the previous blocks and performs a down-mixing, filtering and decimation of each detected signal. The samples for each signal are passed on as a message.
Adapter block to extract the samples of one signal out of the messages from the Signal Separator. Output is again a complex stream.
Resampling can be used to assure a constant output sample rate of the stream. (Needed for FM demod for instance.)
Estimates the parameters subcarrier spacing, symbol time, FFT size and cyclic prefix length for a input signal. The signal should not be oversampled for this block.
After OFDM parameter estimation, the signal is frequency synced and symbol beginnings are marked with stream tags.
Visualisation of FAM data from gr-specest.
Loading of TensorFlow models for AMC.
This GNU Radio module is part of the Google Summer of Code (GSoC) program 2016. The target is to develop a signal analysis / signal intelligence toolbox with the following capabilities:
- Automatic signal detection
- Automatic modulation classification
- OFDM parameter estimation and synchronization
- GUI feedback
- Doxygen documentation
Read the full proposal here.
This project was developed in cooperation with the Communications Engineering Lab of the Karlsruhe Institute of Technology.
The modulation classification is developed by @chrisruk during ESA Summer of Code in Space program.
This software is Copyright © 2016 Free Software Foundation, Inc. and part of GNU Radio. It is free software, and is released under the GPLv3 License.
If you run into problems installing or using gr-inspector, the GNU Radio mailing list is the right place to get help. There are many people who will provide assistance. Please do provide a complete problem description containing
- What exactly where you trying to do?
- What steps did you perform to achieve this?
- What happend (with complete verbose output)?
- Why do you think something went wrong? What should have happend instead?
Visit How to report errors for more detailed instructions.
Please only use GitHub issues for actual bugs and improvements! "This command is not working on my machine" is not a bug.
Maintainer of this module:
Sebastian Müller
gsenpo[at]gmail.com
Christopher Richardson
chrisrichardsonuk[at]gmail.com
If you want to know more about the development process, please visit the blog.