Charu Aggarwal*, Joel Wolf*† and Philip S. Yu*†, *Massachusetts Institute of Technology, Cambridge, Massachusetts; †IBM T.J. Watson Research Center, Yorktown Heights, New York charu at mit.edu, jlw at watson.ibm.com, psyu at watson.ibm.com
A critical issue in the performance of a video-on-demand system is the I/O bandwidth required in order to satisfy client requests. A number of techniques have been proposed in order to reduce these bandwidth requirements. In this paper we concentrate on one such technique, known as adaptive piggybacking. We develop and analyze piggyback merging policies which are optimal over large classes of reasonable methods.