SonicMesh

Acoustic Ultrasonic Data Transfer Library (Research Project)

SonicMesh is a Python library designed for high-frequency ultrasonic communication, enabling data transfer over audio. This project is part of ongoing research into ultrasonic FSK (frequency shift keying) communication and aims to push the limits of audio based data transmission particularly for file transfer tht includes images.

Features

• Send text and files over sound using ultrasonic frequencies
• 64-FSK encoder for efficient data transmission
• FFT-based ultrasonic decoder for accurate reception.
• WAV utils to save and read transmissions
• Exoposes high-level APIs for quick experimentation

Goals

• Enable high speed audio-based transfer of data (images and text for now).
• Explore novel encoding strategies for ultrasonic communication.
• provides a flexible library for research and experimentation

Installation

pip install sonicmesh

Quick Example

from sonicmesh import transmit, decode_wav

#Encoding a message and transmitting over the speaker
transmit("Hello World!!")

# Decoding received signal (from WAV file for now - microphones later)
decoded = decode_wav("message.wav")
print(decoded)

Research Focus

SonicMesh is serious research project aiming to pushing the limits of acoustic communication where users can:

• Experiment with ultrasonic-FSK transmissionn
• Test basic audio-based file transfer, including images (altho its still underdeveloped since the FSK decoding is still not optimized).
• Contribute to development of high-frequency data transmission techniques

Architecture Overview

SonicMesh internally consists of three major components:

1. Encoder

• Convert raw bytes/text into symbols
• Maps each symbol to one of 64 frequencies (64-FSK)
• Generates the final audio waveform for transmission

2. Decoder

• Performs FFT based frequency detection
• extracts symbol sequences from the spectrogram
• converts them back into bytes, text, or file data

3. Acoustic Configuration Defines:

• Sample Rate
• Symbol Duration
• Frequency Table
• Bit depth per symbol (64-FSK -> 6 bits/ sybmol)

Roadmap

Planned areas of development:

• High speed FSK enhancemeents for faster JPEG (and other file) transmission
• Better noise robustness using windowing and adaptive thresholding
• Microphone live decoding (real-time RX path)
• Spectrogram visualization tools for debugging
• Higher-order modulation (128-FSK or chirp based systems)
• Erorr correction codes (Hamming, BCH, or even Reed-Solomon)

Current Limitations

To set the correct expectations:

• File transfer works but is slow at the moment
• FFT decoding still needs a lot of optimization and noise filtration
• Microphone live receive is still in experimental / "in-progress" stage.

Contributing

Contributions are welcome especially inL

• Audio signal processing
• optimizing fsk encoding/decoding
• increasing data transfer speed

All contributions should maintain the research oriented and experiment nature of the project.

License

This project is licensed under the MIT License.