The robotic tape library and optical jukebox provide huge and cheap capacity for the storage of multimedia objects. The stored objects may be retrieved using staging, time slicing, or pipelining. The staging method retrieves the whole objects to the staging buffers prior to consumption at the cost of high start-up latency. The time slice method reduces the start-up latency at the cost of heavy switching overheads. The pipelining methods aim at minimizing the start-up latency. In the normal pipelining method, the sizes of the slices are minimized to maximize the overlapping between the displaying time and the retrieval time of the slices. In space efficient pipelining methods, the buffer size in accessing the slices is minimized. We have already described the normal pipelining and the space efficient pipelining methods in the two previous chapters. The segmented pipelining method to reduce the latency in serving interactive requests is presented in this chapter. Multimedia objects are usually displayed from the beginning to the end in video on demand systems. Interactive video-on-demand systems support VCR-like functions, including fast forward, rewind, pause, and resume functions. Large video systems store many objects. The video systems should allow some searching to allow users find the desired objects. When searching is required, the video-on-demand system would need to provide browsing, jump, keyword, and content based searching. Unless the staging method is used, the multimedia storage system cannot support any VCR-like operations. The segmented pipelining method is designed to provide efficient retrieval of multimedia objects with supporting of previews. In this chapter, the segmented pipelining method is first described in the next section. The performance of the segmented pipelining method is then described and analyzed before this chapter is summarized.
Guaranteeing quality of service in interactive video-on-demand servers
