Please use this identifier to cite or link to this item:
|Title:||Clustering-Driven Wireless Data Broadcasting|
|Authors:||Liaskos, Christos K.|
Papadimitriou, Georgios I.
Pomportsis, Andreas S.
Obaidat, Mohammad S.
|Subjects:||FRASCATI::Natural sciences::Computer and information sciences|
|Source:||IEEE Wireless Communications|
|Abstract:||The performance of a push-based system relies heavily on the proper scheduling of the broadcast data. To this end, the Broadcast Disks method is most commonly employed. It defines a procedure consisting of four separate algorithms: one to provide and handle the clients' feedback, another to group the data objects into disks, a third one to define their spinning velocities, and finally a Broadcast Sequence constructor algorithm. In this article we introduce and evaluate Clustering-Driven Wireless Data Broadcasting (CWDB), a complete instantiation of the Broadcast Disks method. The proposed CWDB procedure addresses the major omissions of preceding schemes, including the total lack of feedback mechanisms, extremely limited variety of spinning velocity definition algorithms, and no thorough and realistic testing of complete combinations of algorithms under various client configurations. A new efficient clustering-driven data grouping algorithm is also introduced. Extensive simulation results are presented, which indicate that the proposed CWDB is absolutely dominant over other classical methods in the vast majority of the test cases and led to very important conclusions that were previously overlooked.|
|Appears in Collections:||Department of Applied Informatics |
Files in This Item:
|WCM-2009-ClusteringDrivenWDB.pdf||528,02 kB||Adobe PDF||View/Open|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.