Home

Awesome

YAFS logo

NEWS

A new version of YAFS is available in the branch.

DESCRIPTION

YAFS (Yet Another Fog Simulator) is a simulator tool based on Python of architectures such as: Fog Computing ecosystems for several analysis regarding with the placement of resources, cost deployment, network design, ... IoT environments are the most evident fact of this type of architecture.

The highlights points of YAFS are:

YAFS is released under the MIT License. However, we would like to know in which project or publication have you used or mentioned YAFS.

Please consider use this cite when you use YAFS:

    Isaac Lera, Carlos Guerrero, Carlos Juiz. YAFS: A simulator for IoT scenarios in fog computing. IEEE Access. Vol. 7(1), pages 91745-91758,
    10.1109/ACCESS.2019.2927895, Jul 10 2019.

    https://ieeexplore.ieee.org/document/8758823

    @ARTICLE{8758823,
    author={I. {Lera} and C. {Guerrero} and C. {Juiz}},
    journal={IEEE Access},
    title={YAFS: A Simulator for IoT Scenarios in Fog Computing},
    year={2019},
    volume={7},
    number={},
    pages={91745-91758},
    keywords={Relays;Large scale integration;Wireless communication;OFDM;Interference cancellation;Channel estimation;Real-time systems;Complex networks;fog computing;Internet of Things;simulator},
    doi={10.1109/ACCESS.2019.2927895},
    ISSN={2169-3536},
    month={},
    }

Resources

YAFS tutorial (https://yafs.readthedocs.io/en/latest/introduction/index.html) and user guide (https://www.slideshare.net/wisaaco/yet-another-fog-simulator-yafs-user-guide) are a good starting point for you. You can also try out some of the Examples (https://yafs.readthedocs.io/en/latest/examples/index.html) shipped with YAFS but in any case you have to understand the main concepts of Cloud Computing and other related architectures to design and modelling your own model.

Installation

NOTE A full implementation in Python 3.6 is available in other repository without any of the example folders repo url. A total adaptation of the example is required to update this repo. Sorry for that, but the working is still the same!

YAFS requires Python 2.7 (Python 3.6 or above is not supported)

  1. Clone the project in your local folder:
    $ git clone https://github.com/acsicuib/YAFS
  1. Install third-libraries with easy_install or pip commands
   $ conda env update -f yafs.yml

Getting started & your first execution

To run some folder project you can create a simple bash script, with the following lines (please update the path according with your system) or you can use a python editor such as: Pycharm, Spyder, etc.

export PYTHONPATH=$PYTHONPATH:/<your path>/YAFS/src/:src/examples/Tutorial/
python src/examples/Tutorial/main1.py

The YAFS tutorial is a good starting point for you. You can also try out some of the Examples shipped with YAFS but in any case you have to understand the main concepts of Cloud Computing and other related architectures to design and modelling your own model.

A "SUPER" TIP: creating custom strategies

We try to implement a wonderful tutorial but our time is limited. Thus, we can introduce this simple tip to create a custom strategy in the simulation. This example defines a dynamical strategy since along the simulation can altered other strategies or even the topology, the applications, or whatever thing that you wish to modify.

For example, in your main.py function, you can declare a custom strategy with a deterministic distribution, and you can include the parameters that you want, i.e. the simulator class, and the routingPath.

dStart = deterministicDistributionStartPoint(400, 100, name="Deterministic")
evol= CustomStrategy()
s.deploy_monitor("EvolutionOfServices",evol,dStart,**{"sim":s,"routing":selectorPath})

And finally you define the CustomStrategy() class:

class CustomStrategy():

    def __call__(self, sim,routing):
        sim.print_debug_assignaments()
        routing.print_control_services()
        routing.my_var = False
        #or whatever you want

Note: the sim variable is type of core.py, it means, it contains all the variables and strategies: topology, placements, applications, etc.

Graph animations

As you can implement events (custom strategies), you can generate plots of your network in each event. Thus, you can store png files and at the end of your simulation, you generate a video with the combination all of them using ffmpeg command.

You can find some examples in the following src/examples: DynamicWorkloads, ConquestService, and mobileTutorial. From DynamicWorkload folder and ConquestService, we have generate the following animations:

ffmpeg -r 1 -i net_%03d.png -c:v libx264 -vf fps=1 -pix_fmt yuv420p out.mp4
ffmpeg -i out.mp4 -pix_fmt rgb24  out.gif

from example/DynamicWorkload <img src="https://github.com/acsicuib/YAFS/raw/master/src/examples/DynamicWorkload/figure/out.gif" width="350" height="350"/></a>

from example/ConquestService <img src="https://github.com/acsicuib/YAFS/raw/master/src/examples/ConquestService/out.gif" width="350" height="350"/></a>

from example/mobileTutorial <img src="https://github.com/acsicuib/YAFS/raw/master/src/examples/mobileTutorial/exp/results_20190326/out.gif" width="550" height="550"/></a>

from example/mobileTutorial <img src="https://github.com/acsicuib/YAFS/raw/master/src/examples/mobileTutorial/exp/results_20190326/out2.gif" width="550" height="550"/></a>

Documentation and Help

The documentation contains a tutorial, the architecture design explaining key concepts, a number of examples and the API reference.

For more help, contact with the authors or You must dig through the source code

Improvements

Acknowledgment

Authors acknowledge financial support through grant project ORDCOT with number TIN2017-88547-P (AEI/FEDER, UE)

REFERENCES

YAFS is used in the following projects:

Please, send us your reference to publish it!