ABR-DQN
Adaptive bitrate (ABR) video streaming is one of the main applications for providing higher video services on all network conditions. In this work, we use Deep Q-learning method to genreate an ABR alogrithm. Meanwhile, we package the Pensieve environments to a simple gym. Results show that ABR algorithm which DQN generates also performs well in various network conditions.
In this sample, please type 'python main.py' to run and enjoy.