Minimizing maximum response time in scheduling broadcasts

We study novel scheduling problems that arise when a data server broadcasts data items requested by users. Two or more users may have pending requests for the same data item at any given time - such `common' requests may be satisfied by broadcasting that data once. Hence, in contrast to the classical scheduling scenario, the scheduling problems arising in the broadcast case have the property that the work done for one job may satisfy some of the others. We formulate the broadcast data delivery model and the scheduling problems that arise. We present polynomial time, near-optimal, offline algorithms as well as tight bounds on the competitive ratio of the online algorithms for the basic problem of minimizing the maximum response time of a request.