netlab-lcy/DRL-OR-S

Code for DRL-OR-S on TPDS 2023

Scalable Deep Reinforcement Learning-Based Online Routing for Multi-type Service Requirements

This is a Pytorch implementation of DRL-OR-S on TPDS 2023.

Requirements

To run DRL-OR-S

• Firstly you should install mininet

git clone git://github.com/mininet/mininet
cd mininet
util/install.sh -a

• Then install ryu-manager and other required packages

pip3 install ryu
pip3 install -r requirements.txt

Offline training DRL-OR-S with NN-simulator

We have uploaded the training and testing data for NN-simulator.

To training DRL-OR-S offline

• Run NN-simulator

cd NN-simulator
python3 main.py --mode simulator --log-dir final-loss-100queuesize --use-cuda --topo Abi --simu-port 5010

• Run DRL-OR-S algorithm

python3 main.py --mode train --use-gae --num-mini-batch 1 --use-linear-lr-decay --num-env-steps 3000000 --env-name Abi --log-dir ./log/test --model-save-path ./model/test --model-load-path ./model/Abi-heavyload-gcn-sharepolicy-SPRsafe-mininet-2penalty-test-Abi-dynamic-extra2   --num-pretrain-epochs 0 --simu-port 5010

Online testing DRL-OR-S with mininet testbed

To run DRL-OR code as an example

• Run testbed

cd testbed
sudo ./run.sh

• Run ryu controller

cd ryu-controller
./run.sh

• Run DRL-OR-S algorithm

cd drl-or-s
./run.sh

If you have any questions, please post an issue or send an email to chenyiliu9@gmail.com.

Citation

@article{liu2023scalable,
  title={Scalable Deep Reinforcement Learning-Based Online Routing for Multi-Type Service Requirements},
  author={Liu, Chenyi and Wu, Pingfei and Xu, Mingwei and Yang, Yuan and Geng, Nan},
  journal={IEEE Transactions on Parallel and Distributed Systems},
  year={2023},
  publisher={IEEE}
}