The gr-lora project aims to provide a collection of GNU Radio blocks for receiving LoRa modulated radio messages using a Software Defined Radio (SDR). More information about LoRa itself can be found on the website of the LoRa Alliance.
All features of the LoRa physical-layer modulation scheme are described in various patents and blog posts (for a good resource, see this RevSpace page). gr-lora supports most of these features, except for:
- CRC checks of the payload and header
- Decoding multiple channels simultaneously
This library was primarily tested with a USRP B201 as receiver and Microchip RN2483 as transmitter. If you encounter an issue with your particular setup, feel free to let me know in the 'Issues' section of this repository.
I'm happy to announce that as of gr-lora version 0.6, a new clock recovery algorithm has been implemented which fixes previous issues with long LoRa messages. Other components, such as whitening, detection and decoding have been improved as well (see the Git log for more details). Given a clear signal, the decoding accuracy is now close to 100% for all SFs, and I therefore consider LoRa fully reverse engineered. Future updates will focus on improving the performance and minor details of the specification.
If you are working on a research topic or project that involves the usage of gr-lora or its algorithms, we would appreciate it if you could acknowledge us. We are currently working on a full writeup of the decoder, but in the meantime, you can cite this repository as follows:
Pieter Robyns, Peter Quax, Wim Lamotte, William Thenaers. (2017). gr-lora: An efficient LoRa decoder for GNU Radio. Zenodo. 10.5281/zenodo.853201
Installing gr-lora is possible in two ways: either by downloading the Docker container, which contains all dependencies and gr-lora packaged in a single container, or by manual installation.
To avoid installation errors and cluttering your system with the required dependencies, the best approach to install gr-lora is through a Docker container. Make sure the docker service is running, then perform the following steps:
git clone https://github.com/rpp0/gr-lora.git .
cd docker/
./docker_run_grlora.sh
The docker_run_grlora.sh script will download the Docker container and run it in 'privileged' mode (in order to access your SDR over USB). After that, you should get a shell inside the container:
[root@5773ed19d95d apps]#
See the 'Testing' section below for examples on how to use gr-lora.
If you prefer a manual installation, the following dependencies are required: python2-numpy, python2-scipy, swig, cppunit, fftw, gnuradio, libvolk, log4cpp, cmake, wx, and liquid-dsp.
The installation procedure is the same as for any GNU Radio OOT module:
git clone https://github.com/rpp0/gr-lora.git .
mkdir build
cd build
cmake ../ # Note to Arch Linux users: add "-DCMAKE_INSTALL_PREFIX=/usr"
make && sudo make install
To test your installation, you can simply run the example app apps/lora_receive_file_nogui.py. The script will download an example trace, and attempt to decode it using gr-lora. You should see the following output:
$ ./lora_receive_file_nogui.py
[?] Download test LoRa signal to decode? [y/N] y
[+] Downloading https://research.edm.uhasselt.be/probyns/lora/usrp-868.1-sf7-cr4-bw125-crc-0.sigmf-data -> ./example-trace.sigmf-data . . . . . . . . . . . . . . . . . .
[+] Downloading https://research.edm.uhasselt.be/probyns/lora/usrp-868.1-sf7-cr4-bw125-crc-0.sigmf-meta -> ./example-trace.sigmf-meta . .
[+] Configuration: 868.1 MHz, SF 7, CR 4/8, BW 125 kHz, prlen 8, crc on, implicit off
[+] Decoding. You should see a header, followed by 'deadbeef' and a CRC 5 times.
Bits (nominal) per symbol: 3.5
Bins per symbol: 128
Samples per symbol: 1024
Decimation: 8
04 90 40 de ad be ef 70 0d
04 90 40 de ad be ef 70 0d
04 90 40 de ad be ef 70 0d
04 90 40 de ad be ef 70 0d
04 90 40 de ad be ef 70 0d
[+] Done
Other example traces can be found in the gr-lora-samples repository.
If you have a hardware LoRa transmitter, you use apps/lora_receive_realtime.py to decode signals in real time. With a Microchip RN2483, you can use python-loranode to easily send messages via Python.
By default, decoded messages will be printed to the console output. However, you can use a message_socket_sink to forward messages to port 40868 over UDP. See the tutorial for more information.
Contributions to the project are very much appreciated! If you have an idea for improvement or noticed a bug, feel free to submit an issue. If you're up for the challenge and would like to introduce a feature yourself, we kindly invite you to submit a pull request.
The following LoRa modules and SDRs were tested and work with gr-lora:
Transmitters: Pycom LoPy, Dragino LoRa Raspberry Pi HAT, Adafruit Feather 32u4, Microchip RN 2483 (custom board), SX1276(Custom Board with STM32 Support) Receivers: HackRF One, USRP B201, RTL-SDR, LimeSDR(LMS7002M)-LimeSDR USB.
- Version 0.7 : Added support for downlink signals, reduced rate mode when using implicit header and arbitrary bandwidths (experimental).
- Version 0.6.2: Improved Message Socket Sink and higher sensitivity to low-power signals.
- Version 0.6.1: Minor bug fixes and improvements.
- Version 0.6 : Significantly increased decoding accuracy and clock drift correction.
- Version 0.5 : Major overhaul of preamble detection and upchirp syncing
- Version 0.4 : Support for all spreading factors, though SFs 11 and 12 are still slow / experimental
- Version 0.3 : Support for all coding rates
- Version 0.2.1: Fixed some issues reported by reletreby
- Version 0.2 : C++ realtime decoder, manual finetuning for correcting frequency offsets of the transmitter.
- Version 0.1 : Python prototype file based decoder, SF7, CR4/8
See the LICENSE file and top of the source files for the license of this project.