High-dimensional Video Caching Selection Method Based on Deep Reinforcement Learning

Abstract

At present, the quantity and the quality of the videos are increasing, which results in more and more video traffic. The combination of video caching and edge computing can improve the performance of multimedia services system. However, the popularity of the videos changes over time, so video caching selection has the dynamic characteristic. Besides, the edge server needs to select some videos from a large number of videos for caching, so video caching selection has the high-dimensional characteristic. In order to reduce the time delay and the traffic cost, we propose a high-dimensional video caching selection method based on deep reinforcement learning. First, the system model of the video caching selection problem is constructed. Second, the video caching action is selected based on the improved deep deterministic policy gradient. Finally, simulation results show that the proposed method can further reduce the time delay of video transmission and the traffic cost of user expense, compared with similar methods.