Abstract
This thesis considers environments wherein the crographics. bandwidth available to a server is insufficient to serve the clients one at a time. In such environments, broadcasting data to all clients is efficient. This thesis deals with issues related to on-demand data broadcasting. We look at the problem of data broadcasting in an environment where clients make explicit requests to the server. The server broadcasts requested data items to all the clients, including those who have not requested the item. This thesis evaluates two new broadcast scheduling algorithms for such an on-demand model with the objective of minimizing the waiting time at clients. A new caching scheme for clients is also proposed. This caching scheme uses information from the server to make caching decisions. A hierarchical model for data broadcasting, using proxy servers between the clients and the main server, is also evaluated. Finally, we address the issue of scheduling broadcasts such that the variance of the waiting time is reduced. The proposed algorithm tries to trade-of the variance with the mean access time.
Kothandaraman, Kannan (1998). On-demand data broadcasting. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1998 -THESIS -K68.