Dual-Engine Intelligent Caching: A Joint Optimization Framework for 360° Mobile VR Video Edge Caching

Immersive communication is regarded as a key driver in the evolution of Virtual Reality technologies. To offer immersive experiences, 360° video edge caching has become an effective solution for minimizing latency to popular content. In order to find the optimal placement for 360° video caching, numerous conventional optimization approaches or deep learning methods have been proposed. However, conventional optimization is computationally expensive and unsuitable for online decision-making, while relying solely on deep learning methods can not guarantee the optimality and feasibility of the solutions, such as whether solutions satisfy the constraints of the original caching problem. In this paper, a dual-engine intelligent caching framework that combines operations research and deep reinforcement learning is proposed for 360° video edge caching and content prefetching. This framework introduces a shared hierarchical caching architecture and formulates a global shared caching placement problem. Adopting a cutting-edge pruning approach, the local cache replacement algorithm is proposed. This utilizes deep reinforcement learning to predict cache boundaries, which are then used to prune the search space of integer linear programming efficiently. Numerical results show that the proposed scheme provides significantly improved performance relative to alternative caching schemes.