Sampling, Communication, and Prediction Co-Design for Synchronizing the Real-World Device and Digital Model in Metaverse

Our design KC-TD3 is based on Twin Delayed Deep Deterministic Policy Gradients (TD3) (code).

Method is tested in OpenAI gym. Networks are trained using PyTorch 1.12 and Python 3.8.

We will release the code soon.