Skip to content

Latest commit

 

History

History
5 lines (3 loc) · 625 Bytes

README.md

File metadata and controls

5 lines (3 loc) · 625 Bytes

  • Here we have coded a state-of-the-art mobility management algorithms for 5G networks using Deep Reinforcement Learning (BCQ, DQN, DDPG) implemented by Python Torch and TensorFlow with 20% improvement in the overall handover KPIs;

  • The developed algorithm uses generative AI to construct synthetic experience replay data and train itself on this data in an offline fashion, resulting in a significant reduction in the online training time.

  • The code requires communication with the simulator of Nokia Bell Labs which emulates a near-real-world environment for mobility management and other network layer design tasks.