A novel video delivery mechanism for caching-enabled networks

  • PDF / 1,676,854 Bytes
  • 15 Pages / 439.642 x 666.49 pts Page_size
  • 31 Downloads / 204 Views

DOWNLOAD

REPORT


A novel video delivery mechanism for caching-enabled networks Zhimin Qi1 Received: 8 July 2019 / Revised: 20 April 2020 / Accepted: 11 June 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract The caching-enabled networks, especially Information-Centric Networking (ICN), have attracted much attention from some global research communities, and the corresponding achievements have been highlighted in the field of video delivery. This paper studies more effective video delivery mechanism in order to guarantee better network performance based on two conceptions, i.e., the intermediate caching of popular video segments and the light delivery of identity object. At first, we evaluate the popularity under the dynamic environment, including descending modelling, ascending modelling and period modelling. Then, we propose identity-based light video delivery mechanism; in particular, we introduce Data Plane Development Kit (DPDK) to ensure that all video segments have the same size and to accelerate video transmission by bypassing the kernel. Finally, the simulation experiments are made based on the real YouTube dataset over CERNET network topology, and the results demonstrate that the proposed video delivery mechanism outperforms two thestate-of-the-art mechanisms in terms of cache hit ratio, routing hop count, delivery delay and network load. Keywords Caching-enabled networks · Video delivery · Popularity modelling · Light mechanism · DPDK

1 Introduction Recently, the characteristic of in-network caching has been accepted and canonized by some prevalent networking paradigms, such as Peer-to-Peer Networking (P2PN) [1], Content Delivery Networks (CDN) [2], Information-Centric Networking (ICN) [3], Hybrid ICN (HICN) [4] and IPv6 Content Networking (6CN) [5]. Among them, (i) P2PN, CDN and ICN are proposed by the academic communities, while HICN and 6CN are proposed by the industrial community, i.e., Cisco; (ii) P2PN, CDN and 6CN are overlay solutions, while  Zhimin Qi

[email protected] 1

School of Media, Jilin Normal University, Siping, 136000, China

Multimedia Tools and Applications

ICN and HICN are the built network-layer solutions; (iii) CDN has the most mature and extensive market application (e.g., Akamai [6]), while ICN has the strongest academic value (e.g., ACM international conference on ICN [7]). Under such context, more and more content providers have been emerging, such as YouTube, NetFlix, iQIYI and Tencent, and they accelerate the delivery of videos to edge users by leveraging in-network caching. In spite of this, the current video delivery schemes have some obvious limitations, such as redundant delivery, high network load and long delivery delay. Therefore, the more effective video delivery mechanism for caching-enabled networks should be investigated in order to guarantee better network performance. According to the report in [8], we know that 83% watching time concentrates on 1% videos, which indicates that the popular (or hot) videos only account for a very small per