Distributed prefetching scheme for random seek support in peer-to-peer streaming applications

Author(s):  
Changxi Zheng ◽  
Guobin Shen ◽  
Shipeng Li
Author(s):  
Majed Alhaisoni ◽  
Antonio Liotta

Media streaming is an essential element of many applications, including the emerging area of mobile systems and services. Internet broadcasting, conferencing, video-on-demand, online gaming, and a variety of other time-constrained applications are gaining significant momentum. Yet, streaming in a pervasive environment is not mature enough to address challenges such as scalability, heterogeneity, and latency. In a client-server system, streaming servers introduce computational and network bottlenecks affecting the scalability of the system and mobile client exhibit intermittent behavior and high-latency connections. This chapter explores ways that several proposed peer-to-peer (P2P) streaming systems deploy to address some of these challenges. An initial introduction on P2P network fundamentals and classifications provides the necessary background information to focus on and assimilate the different mechanisms that enable scalable and resilient streaming in a pervasive environment. The most interesting developments are presented in an accessible way by revisiting the features of common P2P streaming applications. This approach helps in identifying a range of burning research issues that are still undergoing investigation.


Author(s):  
Jian Feng ◽  
Kwok-Tung Lo

Video broadcasting is one of the efficient methods in implementing large-scale video-on-demand (VoD) systems. Managing large number of concurrent channels in a video server is the main barrier to implement Harmonic Broadcast (HB) in VoD applications. In this chapter, we propose a new transmission protocol denoted as Harmonic-Based Peer-to-Peer Broadcasting (HPB) to release the burden of HB by dispersing a large portion of workload over a number of clients denoted peer servers. As the bandwidth requirement of each segment in HB is different, we first develop a segment placement algorithm to balance the workload among peer servers. We then study the dynamic nature of peer-to-peer (P2P) paradigm on the design of the system. An analytical model is developed to determine the optimal number of peer servers required for the system under certain level of availability such that the workload of the central server can reach a target level. The results show that the workload of central server can be significantly reduced.


Author(s):  
Erdinc Korpeoglu ◽  
Cetin Sahin ◽  
Divyakant Agrawal ◽  
Amr El Abbadi ◽  
Takeo Hosomi ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document