This project is done with association of Turku University of Applied Sciences.
This package allows end-users to run a sniffing node in their WIFI-networks to estimate how many people are connected to their network based on connected devices. This script counts unique mac addresses connected to the network and publishes a MQTT message with the result. We realize router firmware has tools already in place to see the amount of devices connected, but this allows a more accurate count to be done based on location, and information can easily be displayed to users on the network. An example would be running this proccess on a library network, and a user can see on their main page how 'busy' the library is.
- @Nicholas Navarro - Student at Turku AMK
- @Michalis Iona - Student at Turku AMK
- @Georgi Georgiev - Student at Turku AMK
- Linux Distro that Supports ARP Commands
- Dual-WIFI or Ethernet + WIFI Connection Supported.
- Docker Service Installed and Enabled
- Python 2.7 Virtual Environment
This project was built using a Raspberry Pi 3+ 4GB with 32-bit Kali Linux Installed.
First install python2.7 and Virtual-Environment on your remote node (e.g. Raspberry Pi 3)
If you are running on a Debian-based distro, use the following:
sudo apt-get install virtualenv
sudo apt-get install python2.7
Secondly, get your docker.service started on your device that will be running the mosquitto-mqtt container. This needs to be separate to your raspberry pi, but connected on a network that they both share. For example, A Laptop and Raspberry Pi connected through ethernet on a shared router or network switch.
To start docker, enable using:
sudo systemctl docker.service start
Clone this repository into a directory of your choosing, on both your node and server host
sudo git clone https......
Create a virtual enviroment folder in the same directory using this command:
virtualenv --python=2.7 venv
Then activate this virtual environment by doing the following:
source venv/bin/activate
Now install requirements:
pip install -r requirements.txt
Now your node is ready to be launched. Before deploying this into your target network. First make sure your node is connected to the target wifi, and your device running the mqtt container has the ability to open ports and is visible on the network. A quick way to confirm to to run: ifconfig -a on a linux, take note of your public ip address, and run the following command on your node:
sudo arp-scan [DEVICE IP ADDRESS]
You should get a response like the following:
Run:
sudo apt-get install -y arp-scan`
By default, Kali-linux distributions come with the necessary packages required to run arp-scan. If you are running a different distribution, monitor-mode needs to be supported on your node's WIFI card in order for it to carry out arp-scan commands.
Custom firmware can be applied to unsupported wlan chips on certain rasberry pi's, for further information, please visit:
https://github.com/seemoo-lab/nexmon
Run the script by typing:
python main.py
Flags can be used in combination with main.py. Table of flags and their defaults are listed as:
| Flag | Default | Meaning |
|---|---|---|
| --host | 'localhost' | Host IP |
| --port | '1883' | Port Number |
| --sub | 'home/mac' | MQTT Publish Root |
| --sleeptime | '30' | Time in seconds, reability of the script |
| --roomname | 'Room' | Room name, this is printed in the published message |
Check that docker is running by running:
sudo systemctl status docker.service
Now since docker is running, run this command in the directory of the repository:
docker-compose up -d
The default port is 1883 and the script is written so ports and ip's are configurable. You will need to change the IP ADDRESS variable to your devices IP to ensure published messages are being sent to your mqtt container.
Using MQTT X: https://mqttx.app/ , you can subscribe to the board using hostname: localhost, and port: 1883. The script uses a default path of home/mac. Again this is configurable in the script, and each node can have different boards to publish to, for example: mainlibrary/area1. If you would like to run MQTT.X on a different device altogether, be sure you use the docker container's device IP.
If you wish to host a container and a separate physical location, or you are running the container in a cloud platform
Be sure that port-forwarding is enabled on your system's router if you are running the container on a private network, this will allow communication to the Docker container remotely. When setting up port-forwarding on your router's config page: (Usually 192.168.1.1), be sure to select the device running your docker container as the destination IP. by default 1883 is the configured port for MQTT.
