The kodo-ns3-examples repository demonstrates how to use the Kodo erasure coding library (http://steinwurf.com/kodo/) in various ns-3 examples.
ns-3 (http://nsnam.org) is a discrete-event network simulator, targeted primarily for research and educational use. ns-3 is licensed under the GNU GPLv2 license.
The kodo-ns3-examples repository: https://github.com/steinwurf/kodo-ns3-examples
If you have any questions or suggestions about this library, please contact us at our developer mailing list (hosted at Google Groups):
Important: Before asking any questions, please try to precisely follow the instructions here and read the guidelines of our mailing list!
If you have a technical issue with ns-3 itself, then please ask around on the ns-3 mailing list.
A valid Kodo license is required if you wish to use this project. Please request a license by filling out the license request form.
Kodo is available under a research- and education-friendly license, you can see the details here.
If you try to configure without a valid license, then you will get an error!
Please read our general documentation here to get started: http://docs.steinwurf.com
The kodo-ns3-examples tutorial is located here: http://docs.steinwurf.com/kodo-ns3-examples/index.html
The tutorial provides a comprehensive description of our examples, and it is designed to be a starting point for ns-3 developers that intend to use the Kodo library for a specific application of Random Linear Network Coding (RLNC) within the simulator.
Using Kodo in an ns-3 example
The examples in this repository will be installed to the
subfolder of your ns-3 folder. We will also build the Kodo static libraries
which will be installed in the same folder together with the required header
files. The examples will be linked with the required static libraries.
As a first step, you need ns-3 installed on your development machine. You may find lots of information about this on the ns-3 webpage:
On Ubuntu/Debian, you need to install the following packages:
sudo apt-get install g++ python mercurial git-core
In the following, we will clone ns-3 to the
~/ns-3-dev folder and we
will clone the kodo-ns3-examples to the
You may use different folders, but the two folders must be separate,
i.e. one cannot be the subfolder of the other.
First clone the ns-3 repository (we start from the home folder,
so it will be cloned to
cd ~ hg clone http://code.nsnam.org/ns-3-dev/
This command will download the ns-3 simulator to your computer into
ns-3-dev folder (this may take a few minutes).
Go to this freshly cloned folder:
Our aim is to make the examples compatible with the latest ns-3 revision. If you experience any issues with the latest revision, then you can switch to the latest supported revision (this step is optional):
hg checkout 12799
Our build system automatically tests the examples with the latest supported revision of ns-3. This revision is specified in our buildbot.py script.
Configure the ns-3 project (it is important to also enable the examples):
python waf configure --enable-examples
This will output a whole bunch of information about the enabled modules based on the availability of tools and libraries installed on your computer.
Now we build the ns-3 libraries and examples:
python waf build
Installing the Kodo examples to ns-3
After building ns-3, you can switch to the kodo-ns3-examples repository.
We will clone this repository to the
cd ~ git clone firstname.lastname@example.org:steinwurf/kodo-ns3-examples.git
Configure this project:
cd kodo-ns3-examples python waf configure
waf configure command ensures that all dependencies are downloaded
(by default, waf will create a folder called
store these libraries).
You must have a valid Steinwurf license to download the
kodo dependencies, otherwise you will get a git error when you execute
the configure command!
Now we build the kodo-rlnc static library and we install the examples and all
the required files to the
python waf build install --ns3_path="~/ns-3-dev"
--ns3_path option is used to specify your ns-3 folder (you can change
this if your ns-3 is located elsewhere).
Building the Kodo examples in ns-3
After the install step, you can switch back to your ns-3 folder:
You can follow the normal ns-3 workflow to build our examples. The ns-3 waf
will automatically find the new examples in
python waf build
We have the following examples:
kodo-wired-broadcast: This example demonstrates broadcasting packets with RLNC from a transmitter to N receivers with the same erasure channel.
kodo-wifi-broadcast: This example demonstrates broadcasting packets with RLNC to N receivers over an IEEE 802.11b WiFi channel.
kodo-recoders: This example shows the gain of RLNC with recoding in a 2-hop line network consisting of an encoder, N recoders and a decoder with different erasure rates. Recoding can be turned on or off and the erasure rates can be modified by command-line options.
You can find more details about each example in their respective source files. There you can also check how to change the simulation parameters like the packet-, field- and generation sizes.
You can run the examples with the usual ns-3 run commands:
python waf --run kodo-wired-broadcast python waf --run kodo-wifi-broadcast python waf --run kodo-recoders
Most of the examples will print out how the decoding matrix changes with each combination packet. You will see if a received packet is linearly dependent or not. You will also see when the decoding process is completed and how many transmissions were required.
Adding your own simulation
At this point, you might want to add your own simulation that uses kodo.
It is recommended to create a new program by copying one of the kodo examples.
The examples are installed in
~/ns-3-dev/examples/kodo, so we will go to
that folder and make a copy
kodo-wifi-broadcast.cc to create a new
my-simulation.cc (you can choose any name here):
cd ~/ns-3-dev/examples/kodo cp kodo-wifi-broadcast.cc my-simulation.cc
To build an executable from the
my-simulation.cc source file, we have to
define a new program in
Open this file in your text editor, and add the following lines at the end
(be careful with the indentation since this is a Python script):
obj = bld.create_ns3_program('my-simulation', ['core', 'applications', 'point-to-point', 'point-to-point-layout', 'internet', 'wifi']) obj.source = 'my-simulation.cc' set_properties(obj)
After this change, the ns-3 waf will detect the new example and you will be
able to run it from the
~/ns-3-dev folder as usual:
cd ~/ns-3-dev python waf --run my-simulation
Now you can expand your custom simulation as you like. If you use additional ns-3 modules, then you need to add them in the wscript (most likely, you will get a build error if you are missing a module).
If your simulation has multiple source files (.cc files), then you can add these in the wscript like this:
obj.source = ['my-simulation.cc', 'source2.cc', 'source3.cc']
Warning: If you install the kodo ns-3 examples again with this command:
cd ~/kodo-ns3-examples python waf build install --ns3_path="~/ns-3-dev"
then the example source files and the wscript will be overwritten in
~/ns-3-dev/examples/kodo folder, so it is recommended to create a
backup if you modified any of these files.
You can also create a separate folder for your custom simulation to avoid this
problem. For example, you can create the
folder and copy the
lib folders from