Design of User Behavior-aware Video Chunk Caching Strategy at Network Edge

This paper introduces an intelligent caching decision framework designed for video chunk caching in mobile edge networks. It uniquely integrates individual user behaviors with broader global trends. The framework utilizes a deep reinforcement learning (DRL)-based approach, which is adept at implementing both reactive and proactive caching strategies. Our model primarily analyzes historical viewing data, leveraging insights from both individual and global user interactions to refine caching decisions. A distinctive feature of our model is its chunk scoring mechanism, which evaluates video chunks based on two criteria: their content similarity to the user's requested sequence and their intrinsic value within the overall video.