kehe-fl 0.1.10

A federated learning package for IoT devices and aggregation server communication.

kehe-fl

A federated learning package for IoT devices and aggregation server communication using MQTT.

Features

• Distributed, federated learning orchestration using MQTT
• Device-side and server-side reference implementations
• Asyncio-based for efficient concurrency
• Modular design for custom ML or IoT projects

Quick Start

Installation

pip install kehe-fl

Example Usage

Device Side

import asyncio
from kehe_fl.comms.mqtt_device import MQTTDevice

mqttConnection: MQTTDevice | None = None

async def main():
    global mqttConnection

    mqttConnection = MQTTDevice(broker="192.168.1.193", deviceId="device123")

    mqtt_task = asyncio.create_task(mqttConnection.connect_and_listen())

    await asyncio.gather(mqtt_task)

asyncio.run(main())

Aggregation Server Side

import asyncio
from kehe_fl.comms.mqtt_agg_server import MQTTAggServer

mqttConnection: MQTTAggServer | None = None

async def handleMessaging():
    global mqttConnection
    loop = asyncio.get_running_loop()

    while True:
        if mqttConnection.is_connected and not mqttConnection.working:
            message = await loop.run_in_executor(None, input, "Enter a command to send to the clients: ")
            await mqttConnection.send_command(message)
        else:
            await asyncio.sleep(2)

async def main():
    global mqttConnection

    mqttConnection = MQTTAggServer(broker="localhost")

    mqtt_task = asyncio.create_task(mqttConnection.connect_and_listen())
    input_task = asyncio.create_task(handleMessaging())

    await asyncio.gather(mqtt_task, input_task)

asyncio.run(main())

Adapt

Communication

You can adapt the device and server code to your specific ML or IoT project needs by modifying the MQTTDevice and MQTTAggServer or even MQTTProvider classes. These classes provide a foundation for communication and can be extended with custom logic for model training, data handling, and more.

Machine Learning

You can integrate your preferred machine learning libraries (like TensorFlow, PyTorch, etc.) into the device and server implementations with modifying ModelService. The provided classes can be used to send model updates, receive commands, and manage the training process across devices.

Data Collection

You can implement custom sensor classes in common/<your sensor>.pyand also modify the DataCollectionService for your needs. This allows you to collect and process data from various IoT sensors and devices, which can then be used for training machine learning models.

Project Constants

You can modify the project_constants.py file to change the MQTT topic structure, message formats, and other constants used throughout the package. This allows you to tailor the communication protocol to your specific requirements.