New models and algorithms for throughput maximization in broadcast scheduling

In this paper we consider some basic scheduling questions motivated by query processing that involve accessing resources (such as sensors) to gather data. Clients issue requests for data from resources and the data may be dynamic or changing which imposes temporal constraints on the delivery of the data. A proxy server has to compute a probing schedule for the resources since it can probe a limited number of resources at each time step. Due to overlapping client requests, multiple queries can be answered by probing the resource at a certain time. This leads to problems related to some well-studied broadcast scheduling problems. However, the specific requirements of the applications motivate some generalizations and variants of previously studied metrics for broadcast scheduling. We consider both online and offline versions of these problems and provide new algorithms and results.