Remote Sensor Data Monitoring with Lossless Techniques

This repository provides domain-specific lossless compression algorithms for resource-constrained Arduino microcontroller platforms. It features a delta-Rice coding scheme for temperature data and a regime-based adaptive Rice coding scheme for LDR (Light Dependent Resistor) data.

Features

• Lossless Compression: Exact byte-for-byte reconstruction.
• Resource Efficient: Requires <256 bytes of RAM on an Arduino Uno during encoding.
• Domain-Specific: Exploits structural properties of physical sensor data (temporal coherence, regime-based behavior).
• Cross-platform Decompression: Python backend for restoring and analyzing data.

Structure

• arduino/: Contains the shared RiceCodec C++ files and the sketches for temperature_compression and ldr_compression.
• python/: Contains the decompress.py backend, statistical analysis tools, and unit tests.

Usage

1. Arduino: Upload the desired sketch to your Arduino with an SD card module connected (CS=10, MOSI=11, MISO=12, SCK=13). The encoded data will be saved as .bin files in the /compressed/ directory on the SD card.
2. Python: Use python/decompress.py <file.bin> to convert the binary files back into readable lists or process a whole directory using the script's batch_process_directory functionality.

Tests

Run the unit tests in the python/ directory using:

python -m unittest test_decompress.py