RIOT Application AWS

Overview

In this project, we have created an IoT (Internet of Things) sensor node utilizing the RIOT operating system. This sensor node is specifically engineered for the collection of environmental data, such as temperature and pressure. The collected data is then securely transmitted to an MQTT (Message Queuing Telemetry Transport) broker through secure protocols. The sensor node establishes communication with an RSMB broker, and the configuration of the Mosquitto client bridge enables a connection to AWS EC2. Subsequently, a Python MQTT client subscribed to the topic stores the data in InfluxDB, and the information is visualized through Grafana.

Directory Structure

└── RIOT-Application-AWS/
    ├── RSMB-broker/
    │   └── config.conf
    ├── Python-MQTT-subscriber/
    │   └── mqtt_subscriber.py
    ├── Mosquitto-Client-Bridge/
    │   └── mosquitto.config
    └── RIOT-Sensor-Node/
        ├── Makefile
        └── main.c

Files and Folders

RSMB-Broker

File	Summary
config.conf	► The config.conf file enables MQTT connections, with debug tracing and specific listeners. It also establishes a connection named "local_to_cloud" between sensor node and Mosquitto Client on A8 node on respective ports.

RIOT-Sensor-Node

File	Summary
main.c	► The C program for the IoT sensor node, named "IoTSensors," includes MQTT-SN communication, LPS331AP sensor readings, and a command-line interface (CLI). It periodically measures temperature and pressure, publishes the data to an MQTT broker, and provides status reports via the CLI. The program creates threads for MQTT communication and the main measurement loop.
Makefile	► The Makefile for the IoTSensors application in RIOT OS configures a native board, includes necessary modules for sensor and network functionality, sets up MQTT modules for communication, and defines parameters such as server address, port, and MQTT topics.

Python-MQTT-Subscriber

File	Summary
NodeRed_Flow.json	► The python script which acts as a mqtt subscriber. It subscribes to the mqtt topic "local_to_cloud" and then it further connects to the InfluxDB to save the payload data in the database.

Mosquitto-Client-Bridge

File	Summary
mosquitto.config	► The mosquitto client bridge which act as a bridge between rsmb broker and AWS EC2 mosquitto client.

Getting Started

Installation

Clone the RIOT-Application-AWS repository:

git clone https://github.com/Saim-Akhtar/RIOT-Application-AWS

Change to the project directory:

cd RIOT-Application-AWS

Running RIOT-Application-AWS

Connect to Grenoble SSH Frontend

ssh <login>@grenoble.iot-lab.info

Start Experiment on IoT-Lab Test Bed

Launch an experiment with two M3 nodes and one A8 node Wait for the experiment to reach the "Running" state. I have used 103,104 of M3 nodes and 106 of A8 node as an example. You can use any other node-ids

iotlab-experiment submit -n iot -d 120 -l grenoble,m3,103-104 -l grenoble,a8,106

Setup Border Rounter on one of the M3 Nodes Source RIOT environment

source /opt/riot.source

Build border router firmware for M3 node with baudrate 500000

make ETHOS_BAUDRATE=500000 DEFAULT_CHANNEL=20 BOARD=iotlab-m3 -C RIOT/examples/gnrc_border_router clean all

Flash Border Router Now flash the border router to the first M3 node (m3-1 in this case)

iotlab-node --flash RIOT/examples/gnrc_border_router/bin/iotlab-m3/gnrc_border_router.elf -l grenoble,m3,103

Configure Border Router Network Choose an IPv6 prefix for the grenoble site (e.g., 2001:660:5307:3100::/64 to 2001:660:5307:317f::/64) Configure the network of the border router on m3-103 Setup a tap interface and an IPv6 prefix with ethos_uhcpd.py

sudo ethos_uhcpd.py m3-103 tap0 2001:660:5307:3100::1/64

Setup RSMB Broker and Mosquitto Bridge on A8 Node. In a new terminal connect to the SSH frontend, and login into clone the mqtt_broker and mosquitto bridge configuration files in A8 shared directory. SSH into the A8 node

ssh root@node-a8-106

Check the global IPv6 address of the A8 node and copy it to use in RIOT Sensor Node

ifconfig

Start Rsmb Broker From the A8 shared directory, start the Rsmb broker using config.conf

cd ~/A8
cd RSMB-Broker
./broker_mqtts config.conf

Configure and Start Mosquitto Client Bridge Kill the ports of any existing running mosquitto client Check the existing mosquitto ports by the following command

ps -ef | grep mosquitto

If there is a port in use, kill the port by using the following command

kill <port>

Modify mosquitto.config with the IPv6 address of the EC2 instance

#Paste your AWS Mosquitto MQTT address in the form of <IPv6 address>:<port>
address <IPv6 address of EC2 instance>:1883

#Example: address 2a05:d016:c59:1d18:70d3:bed4:fc5a:68ac:1883

Start Mosquitto service

root@node-a8-3:~/A8/Mosquitto-Client-Bridge/mqtt_bridge# mosquitto -c mosquitto.conf

Setup and Build Flash RIOT Sensor Node Firmware

From new terminal connect again to SSH front end of grenoble site and enter the RIOT-Sensor-Node

cd RIOT-Sensor-Node

Build the firmware for the riot sensor node using A8 node's IPv6 address (copied previously) and tap-id (i.e. 0 in our case)

make DEFAULT_CHANNEL=20 SERVER_ADDR=<IPv6 address> EMCUTE_ID=station0 BOARD=iotlab-m3 -C . clean all

Flash the riot sensor node firmware on an M3 node 104

iotlab-node --flash ./bin/iotlab-m3/IoTSensors.elf -l grenoble,m3,104

Connect to RIOT Sensor Node

Log into the M3 node

nc m3-104 20000

AWS Cloud EC2

Create EC2 instance and assign IPv6 subnet according to the following tutorial : https://aws.amazon.com/blogs/networking-and-content-delivery/introducing-ipv6-only-subnets-and-ec2-instances/
Login to EC2 instance using SSH and install mosquitto client using apt-get:

sudo apt-get update
sudo apt-get install mosquitto
sudo apt-get install mosquitto-clients
sudo apt clean

Check if the mosquitto service is running (It should show broker running):

sudo service mosquitto status

Allow anonymous connections on mosquitto configurations and enable listener 1883

sudo nano /etc/mosquitto/mosquitto.conf

and then add the follow lines in the file

listener 1883
allow_anonymous true

Then restart the mosquitto service:

sudo service mosquitto restart

Install docker on EC2 instance:

sudo snap install docker

Setup the Influxdb container on EC2 instance:

docker run --detach --name influxdb -p 8086:8086 influxdb:2.2.0

Setup the Grafana container on EC2 instance:

docker run -d --name=grafana -p 3000:3000 grafana/grafana

Make sure to have the following ports publicly accessible in security setting of EC2:

Go to InfluxDB using :8086, setup an organization name and create a bucket.
Start running the python mqtt subscriber:

This is how the data will be visible on InfluxDB:
On Grafana add influxdb as data sources, and add details for the influxdb bucket which you created in previous steps.