Intro

Official implementation of the GLOBECOM 2025 paper Robust Bandwidth Estimation for Real-Time Communication with Offline Reinforcement Learning (offline part)

Quick start

• Download training data and configure the environment according to Schaferct's instructions: README_Schaferct

• We mapped the action space, restructured the dataset, and provided data examples in dir: training_dataset_pickle

• Run the training script:

  cd code
  python riql6_ensemble.py
  

Offline eval

1. To run a small evaluation on a small dataset: (download the 24 sessions and modify their path first)

   cd code
   python detail_evaluate_on_24_sessions.py

2. To evaluate the metrics (mse, errorate) over all evaluation dataset:

   cd code
   python evaluate_all.py

The whole evaluate process takes more than 10 hours.

Online eval

For online evaluation, we developed it based on the Pandia platform, and we also sourced the deployment code: Pandia

1. Clone repo and install packets

2. Run the script

   python -m pandia.agent.env_emulator_offline