Natural redundancy based error correction
My algorithms for error correction based / aided by natural redundancy in data. The package considers natural redundancy which exists in sensor data such as telemetry data. The protocol used to transmit the data also gives rise to additional redundancy. The code currently considers only the MAVLink protocol, commonly used for UAV's.
To run tests simply clone, cd into the cloned repo, and run:
python -m pytest
or
python -m pytest --cov-report=html
to run also coverage tests.
- Algo - Implements Error correction algorithm. Subclass the CorrectionAlgorithm interface to create models.
- Data Models - Implements data models to model the structure of data. Subclass DataModel interface to create models.
- Decoders - Implements structure based decoders. Subclass Decoder interface to create decoders.
- Inference - Implements utilities for general purpose inference, such as segmentation.
- Utils - Includes various utilities.
- Protocol Meta - Includes metadata about the protocol used.
- MAVLink Utils - utility for generating new MAVLink dialects. Based on pymavlink.
- scripts - Includes scripts which should be used as "main" scrips to run full simulations.
- examples - Used to hold small examples to showcase the use of some classes.
For questions or suggestions contact me.