SVC-aware Cache Coordination Schemes for Dynamic Adaptive Streaming in Named Data Networking

As today's Internet usage is shifting from end-to-end communication to large-scale content dissemination, Named Data Networking (NDN) has been proposed as a novel network architecture to address the communication paradigm mismatch between applications and the underlying TCP/IP network. Since NDN shares many common grounds with Dynamic Adaptive Streaming (DAS) and Scalable Video Coding (SVC), it has the potential to take full advantages of DAS and SVC. However, traditional SVC-agnostic cache strategies ignore the inherent difference among data chunks in terms of their importance. In this paper, we designed two SVC-aware probabilistic caching strategies, namely SVC-aware Reverse Probabilistic Caching (SRPC) and SVC-aware Linear Probabilistic Caching (SLPC). Instead of using a uniform caching probability, the probability to cache a data chunk varies according to both the layer index of the SVC-encoded video chunk and the content router's location along the path from the client to the server. Simulation results based on ndnSIM showed that our proposed SVC-aware strategies outperform SVC-agnostic ones.