Design and analysis of stream scheduling algorithms in distributed reservation-based multimedia systems

Abstract

Video-on-Reservation (VoR) serves as an attractive service providing scheme for network-based multimedia services. In such a service, clients can request multimedia files and their preferred viewing times in advance to the actual playback. We address the issue of how to distribute continuous video streams over a set of heterogeneous media servers in such a way that the consumed network resources associated with network cache space and server bandwidth can be minimized. We propose source-based (SBS) and destination-based (DBS) stream scheduling algorithms, in which time-constrained multicast trees are constructed to guarantee Quality of Services in a cost-effective way. A window-assisted video partitioning (WAVP) strategy is also proposed to partition videos into multiple portions, cache stream portions dynamically, and replace them according to adaptive schedule windows. We make theoretical analysis and study the performance under various influencing factors such as request rate, time constraint, length of streams portions, caching and transmission price ratios. Our simulation and experimental results show those strategies are able to achieve much lower average service cost, balance the load of servers, and accept more number of requests.