Collaborative Edge Caching for Panoramic Video Streaming

With the popularity of virtual reality applications in daily life, a large number of users begin to acquaint themselves with panoramic videos. However, the demand for high bandwidth and low latency brings challenges to the transmission of panoramic videos. Although the tile-based viewport adaptive streaming scheme can mitigate this problem and improve bandwidth utilization, the sudden viewport switching of the users in an immersive environment will result in high transmission delay and video playback stall, which reduces the user’s quality of experience. Prefetching the content in panoramic videos that users most likely watch to an edge server can further reduce transmission latency. However, due to the limitation of edge server cache capacity, better caching strategies need to be proposed. In this paper, we propose an edge collaborative cache scheme, which performs cache and replacement decisions to improve cache utilization and minimize transmission delay and cost. Moreover, adjacent multiple mobile edge computing (MEC) servers form a collaboration domain and share the cache capacity. The proposed strategy determines the proactive caching content and selects the best cache server by taking the cache benefit and replacement cost of the whole cooperation domain into account. Experimental results show that compared with FIFO and LRU, the proposed cache replacement strategy has a higher cache hit ratio and viewport quality, and lower stall times, which can bring a good user experience.