ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Churn-Aware Optimal Layer Scheduling Scheme for Scalable Video Distribution in Super-Peer Overlay Networks
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Yong-Hyuk Moon, Jeongnyeo Kim, Chan-Hyun Youn
Issue Date
2012-05
Citation
High-Performance Grid and Cloud Computing Workshop (HPGC) 2012, pp.1-5
Language
English
Type
Conference Paper
Abstract
The server-based content delivery system has experienced a large amount of operation costs and also has faced the lack of adaptive control for differentiated streaming services. To address this problem, some works have shown that the possibility of large-scale data dissemination can be achieved over peer-to-peer networks, if the data (particularly video) is constructed by the scalable video coding. Despite the expected strengths of integrating the two techniques, it is still difficult to ensure collective bandwidth (layer) availability in peer-to-peer networks, where large-scalable, volatile, and heterogeneous peers demand different quality of services (QoS). In this paper, we first formulate a layer scheduling problem from understanding some constraints, which originates from the layer dependency, the transmission rule, and the bandwidth heterogeneity. We then propose that this problem can be solved based on the ideas of how a threshold layer index of scalable video can be determined. By differently prioritizing each layer, the proposed layer scheduling scheme can guarantee the differentiated streaming services using a random scheduling algorithm, providing low complexity in decision. To cope with dynamic intensity of churns in super-peer overlay networks, we also propose a peer-utility based promotion (PUP) algorithm that selects the most qualified neighbor to guarantee the sustained quality of streaming. Simulation results show that the proposed layer scheduling scheme with PUP highly outperforms those with different peer selection strategies in terms of the average bandwidth (6.9 % higher at least) and the variation of utilization (11.3 % lower at least) without violating the constraints on streaming video and network properties.
KSP Keywords
Data Dissemination, Delivery systems, Large-scale Data, Low complexity, Network Properties, Overlay networks, Peer selection, Peer-to-Peer(P2P), Random Scheduling, Scalable Video Coding, Scheduling Scheme