ADeLE: v1.0 README
Updated Dec 18, 2018 by Alessandro Gaballo  

All feedback appreciated to alessandro.gaballo@studenti.polito.it

What is ADeLE
================
ADeLE is an architecture that expands the classical notion of computation offloading by identifying the invariances of the edge
offloading problem; in particular, we consider the problem of traffic offloading, i.e., the SDN-driven management mechanism
to route traffic among processes involved in the offloading process.


DISTRIBUTION
================
The distribution tree contains:
  
README.txt
      --this file  
configs
      --the file configurations for MinineXt and Quagga
dataset_final
      --the dataset used to train the LSTM, separated in each generation step  
deep_learning  
      --the Deep Learning models (DNN, LSTM)
evaluation
      --evaluations scripts and results for different metrics
offloading_architecture
      --files for the offloading mechanism and the protocol definition (protobuf subdirectory). 
        These files are a preliminary test to familiarize with ryu and test the protocol.
ryu
      --ryu application to manage the switches and collect data
second_eval
      --additional evaluations
trained_models
      --trained models for both DNN and LSTM
utils
      --utils script
pair_dataset.py
      --builds the dataset by combining the packet counter and the routing table in each run
start.py
      --starts the mininet network for the dataset generation
topology.py
      --mininext topology definition
      
      
RUN
================
To generate the dataset, after setting the parameters in the two scripts to the desired values, run in separate shells:

ryu-manager switchWithStats.py (from within the ryu directory)
sudo python start.py