ABSTRACT

A new architecture for defining network collaboration in DASH is called SAND. It aims to standardize communication amongst network agents participating in the streaming process. This work produces a novel taxonomy for SAND-based approaches. It places AVS solutions into four categories: (1) Management Architecture (2) Cache-based, (3) Optimization, and (4) Paradigm. Sub-categories of Management Architecture are (a) Cognitive, (b) Prioritization, and (c) Encoding and Signaling, while sub-categories of Paradigm are (a) SDN-based, (b) CDN-based, and (c) CCN-based. The approaches representing each category are presented. It is shown that using the SANDs architecture proves beneficial to approaches adopting it. In recent years Cloud-based strategies for AVS have been an important area of research and industry. This paper elaborates on Cloud-based strategies. It places current strategies into a taxonomy. The Cloud-based AVS strategies are categorized into (1) SocialAwareness, (2) Cloud-based CDN, (3) Cloud-based Gaming, (4) Cloud-based SDN, (5) File-hosting services, (6) View synthesis techniques, and (7) Error Concealment techniques. The implementations that exist for the Cloudbased strategies are discussed. Thus, the advantages of both SAND and Cloud-based approaches for use in DASHbased systems are illustrated.

Keywords: - DASH; SAND; Adaptive; video; streaming; SDN; optimization; cache; CDN; Encoding; Signaling; Cloud; Social-Awareness; Gaming.