This paper proposes a novel social-viewport adaptive caching scheme (SACS) for virtual reality (VR) streaming in an edge-computing platform. In VR contents with 360 degree views where only a part of the entire view (i.e., the viewport) is shown and the remaining parts are decoded but not shown, we collect and record multiple clients’ viewports of the same VR contents in local proximity on the edge-computing platform. We extract a social-viewport map, which represents where most of the local clients are directing their attention. By utilizing the social-viewport map, under our proposed scheme, k-means and mean-shift clustering algorithms are adopted to partition 360 degree views into multiple clusters with the nearest mean of hit-ratios from multiple clients. Accordingly, in order to save cache storage while maintaining a high-quality VR streaming service, we adaptively assign different encoding rates with various levels to multiple viewports. We implement the proposed scheme using a commercial EdgeX foundry edge-computing platform. A measurement-based experiment reveals that the proposed scheme achieves a maximum storage reduction of almost 74%, with a 92% hit-ratio to the highest encoded viewports.

本文为边缘计算平台中的虚拟现实（VR）流提出了一种新颖的社交视口自适应缓存方案（SACS）。在具有360度视图的VR内容中，其中仅显示了整个视图（即视口）的一部分，而其余部分被解码但未显示，我们在本地附近收集并记录相同VR内容的多个客户端视口边缘计算平台。我们提取了一个社交视口图，该图代表了大多数本地客户在引导他们的注意力的地方。通过利用社交视口图，在我们提出的方案下，采用k均值和均值漂移聚类算法将360度视图划分为多个聚类，其中多个客户的点击率均值最接近。因此，为了节省高速缓存存储量，同时保持高质量的VR流服务，我们将具有不同级别的不同编码率自适应地分配给多个视口。我们使用商业EdgeX铸造边缘计算平台来实施所提出的方案。基于测量的实验表明，所提出的方案可实现最大存储量减少近74％，与最高编码视口的命中率达到92％。