In recent years, there has been a dramatic increase in the bandwidth of networks, technological advances in the power of central processing units (CPUs) and modern storage systems. The availability of mass communication between them makes it possible to process and deliver media-on-demand (MoD) services to customers in a cost-effective way. Media-on-demand systems is a collection of computers that store the compressed digitized-media sources such as video clips, audio, advertisement, entertainment and distance learning and provide the services to multiple clients simultaneously. The crucial criterion for continuous playback transmission of real-time media data requires a guaranteed bandwidth at a specific rate. One of the technical challenges in the design of a media-on-demand s...[ Read more ]

In recent years, there has been a dramatic increase in the bandwidth of networks, technological advances in the power of central processing units (CPUs) and modern storage systems. The availability of mass communication between them makes it possible to process and deliver media-on-demand (MoD) services to customers in a cost-effective way. Media-on-demand systems is a collection of computers that store the compressed digitized-media sources such as video clips, audio, advertisement, entertainment and distance learning and provide the services to multiple clients simultaneously. The crucial criterion for continuous playback transmission of real-time media data requires a guaranteed bandwidth at a specific rate. One of the technical challenges in the design of a media-on-demand system is the scalability of user population. To cope with the large-scale media-on-demand system, a hierarchical distributed large storage subsystem is required. The overall design of software architecture and network infrastructure in detail is described and emphasis is placed on the underlying server-client model software prototype implementations that can be deployed to support large number of users population. On the server-side, the functionalities, such as request scheduling, data buffering and storage replacement, are investigated. In this research work, a low-cost scalable distributed media-on-demand system is implemented and demonstrated. A focal point of this work is also to demonstrate the use of existing technologies to provide media-on-demand services.