Table of Contents

• RPL Attacks Framework
• System Requirements
• Installation
• Non-Standard Configuration
• Quick Start (using the integrated console)
• Quick Start (using fabric)
• Commands
• Simulation campaign

RPL Attacks Framework

This project is aimed to provide a simple and convenient way to generate simulations and deploy malicious motes for a Wireless Sensor Network (WSN) that uses Routing Protocol for Low-power and lossy devices (RPL) as its network layer.

With this framework, it is possible to easily define campaign of simulations either redefining RPL configuration constants, modifying single lines from the ContikiRPL library or using an own external RPL library. Moreover, experiments in a campaign can be generated either based on a same or a randomized topology for each simulation.

A few test cases made with the framework:

Test case 1: flooding attack

The malicious mote has 3, 7, 10 in its range	Power tracking without the malicious mote	Power tracking with the malicious mote

Test case 2: versioning attack

Legitimate DODAG	Versioning attack in action (global repair)

Power tracking without the malicious mote	Power tracking with the malicious mote

Test case 3a: blackhole attack

Legitimate DODAG	Blackhole attack in action

Test case 3b: blackhole attack

Legitimate DODAG	Blackhole attack in action

System Requirements

This framework was tested on an InstantContiki appliance (that is, an Ubuntu 14.04).

It was tested with Python 2 and 3.

Installation

1. Clone this repository

git clone https://github.com/dhondta/rpl-attacks.git

Behind a proxy ?

Setting: git config --global http.proxy http://[user]:[pwd]@[host]:[port]

Unsetting: git config --global --unset http.proxy

Getting: git config --global --get http.proxy

2. Install system requirements

sudo apt-get install gfortran libopenblas-dev liblapack-dev
sudo apt-get install python-dev libffi-dev libssl-dev libxml2-dev libxslt1-dev libjpeg8-dev zlib1g-dev
sudo apt-get install python-numpy python-scipy
sudo apt-get install imagemagick libcairo2-dev libffi-dev

Behind a proxy ?

Do not forget to configure your Network system settings (or manually edit /etc/apt/apt.conf).

If not using InstantContiki appliance, also install :

sudo apt-get install build-essential binutils-msp430 gcc-msp430 msp430-libc msp430mcu mspdebug binutils-avr gcc-avr gdb-avr avr-libc avrdude openjdk-7-jdk openjdk-7-jre ant libncurses5-dev lib32ncurses5

3. Install Python requirements

sudo apt-get install python-pip
sudo pip install -r requirements.txt

or

sudo apt-get install python3-pip
sudo pip3 install -r requirements.txt

Behind a proxy ?

Do not forget to add option --proxy=http://[user]:[pwd]@[host]:[port] to your pip command.

4. Setup dependencies and test the framework

../rpl-attacks$ fab test

Non-Standard Configuration

This section only applies if you want to tune Contiki's source folder and/or your experiments folder.

Create a default configuration file

../rpl-attacks$ fab config

or create a configuration file with your own parameters (respectively, contiki_folder and experiments_folder)

../rpl-attacks$ fab config:/opt/contiki,~/simulations

Parameters :

• contiki_folder: the path to your contiki installation

[default: ~/contiki]

• experiments_fodler: the path to your experiments folder

[default: ~/Experiments]

These parameters can be later tuned by editing ~/.rpl-attacks.conf. These are written in a section named "RPL Attacks Framework Configuration".

Example configuration file :

[RPL Attacks Framework Configuration]
contiki_folder = /opt/contiki
experiments_folder = ~/simulations

Quick Start (using the integrated console)

1. Open the console (you should see something like in the following screenshot)

../rpl-attacks$ fab console

or

../rpl-attacks$ python main.py

or

../rpl-attacks$ python3 main.py

2. Create a campaign of simulations

user@instant-contiki:rpl-attacks>> prepare sample-attacks

3. Go to your experiments folder (default: ~/Experiments) and edit your new sample-attacks.json to suit your needs

See for more information.

4. Make the simulations

user@instant-contiki:rpl-attacks>> make_all sample-attacks

5. Run the simulations (multi-processed)

user@instant-contiki:rpl-attacks>> run_all sample-attacks

Hint : You can type status during make_all and run_all processing for getting the status of pending tasks.

6. Once tasks are in status SUCCESS in the status tables (visible by typing status), just go to the experiment's results folders to get pictures and logs of the simulations. The related paths are the followings :

[EXPERIMENTS_FOLDER]/[experiment_name]/without-malicious/results/ [EXPERIMENTS_FOLDER]/[experiment_name]/with-malicious/results/

Quick Start (using fabric)

1. Create a simulation campaign file from the template

../rpl-attacks$ fab prepare:test-campaign

2. Edit the simulation campaign file to suit your needs

3. Create the simulations

../rpl-attacks$ fab make_all:test-campaign

4. Run the simulations (not multi-processed)

../rpl-attacks$ fab run_all:test-campaign

5. Once done, just go to the experiment's results folders to get pictures and logs of the simulations. The related paths are the followings :

[EXPERIMENTS_FOLDER]/[experiment_name]/without-malicious/results/ [EXPERIMENTS_FOLDER]/[experiment_name]/with-malicious/results/

Commands

Commands are used by typing fab [command here] (e.g. fab launch:hello-flood) or in the framework's console (e.g. launch hello-flood).

• buildname

This will the malicious mote from the simulation directory named 'name' and upload it to the target hardware.

• cleanname

This will clean the simulation directory named 'name'.

• config[contiki_folder, experiments_folder]

This will create a configuration file with the given parameters at ~/.rpl-attacks.conf.

contiki_folder: path to Contiki installation [default: ~/contiki]

experiments_folder: path to your experiments [default: Experiments]

• coojaname[, with-malicious-mote]

This will open Cooja and load simulation named 'name' in its version with or without the malicious mote.

with-malicious-mote: flag for starting the simulation with/without the malicious mote [default: false]

• dropsimulation-campaign-json-file

This will remove the campaign file named 'simulation-campaign-json-file'.

• listtype-of-item

This will list all existing items of the specified type from the experiment folder.

type-of-item: experiments or campaigns

• makename[, n, ...]

This will create a simulation named 'name' with specified parameters and also build all firmwares from root.c, sensor.c and malicious.c templates with the specified target mote type. This can alternatively make the malicious mote with an external library by providing its path.

n: number of sensors (excluding the root and malicious motes)

duration: simulation duration in seconds

title: simulation title

goal: simulation goal (displayed in the Notes pane)

notes: simulation notes (appended behind the goal in the Notes pane)

min_range: malicious mote's maximum range from the root

tx_range: transmission range

int_range: interference range

area_side: side of the square area of the WSN

mtype_root: root mote type

mtype_sensor: sensor mote type

mtype_malicious: malicious mote type

malicious_target: external RPL library for building the malicious mote

blocks: building blocks for building the malicious mote, as defined in ./templates/building-blocks.json

ext_lib: external RPL library for building the malicious mote

• make_allsimulation-campaign-json-file

This will generate a campaign of simulations from a JSON file.

• preparesimulation-campaign-json-file

This will generate a campaign JSON file from the template located at ./templates/experiments.json.

• remake_allsimulation-campaign-json-file

This will re-generate malicious motes for a campaign of simulations from the selected malicious mote template (which can then be modified to refine only the malicious mote without re-generating the entire campaign).

• runname

This will execute the given simulation, parse log files and generate the results.

• run_allsimulation-campaign-json-file

This will run the entire simulation campaign.

• setup

This will setup Contiki, Cooja and upgrade msp430-gcc for RPL Attacks.

• status

This will show the status of current multi-processed tasks.

• test

This will test the framework.

• update

This will attempt to update Git repositories of Contiki-OS and RPL Attacks Framework.

• versions

This will display the versions of Contiki-OS and RPL Attacks Framework.

Simulation campaign

Example JSON for a campaign with a BASE simulation as a template for the other simulations (with the same topology) :

Example JSON for a campaign of heterogeneous simulations (with randomized topologies) :