The problem of data broadcasting over multi- ple channels consists in partitioning data among channels, depending on data popularities, and then cyclically trans- mitting them over each channel so that the average wait- ing time of the clients is minimized. Such a problem is known to be polynomially time solvable for uniform length data items, while it is computationally intractable for non- uniform length data items. In this paper, two new heuris- tics are proposed which exploit a novel characterization of optimal solutions for the special case of two channels and data items of uniform lengths. Sub-optimal solutions for the most general case of an arbitrary number of channels and data items of non-uniform lengths are provided. The first heuristic, called Greedy +, combines the novel characteri- zation with the known greedy approach, while the second heuristic, called Dlinear, combines the same characteriza- tion with the dynamic programming technique. Such heuris- tics have been tested on benchmarks whose popularities are characterized by Zipf distributions, as well as on a wider set of benchmarks. The experimental tests reveal that Dlinear finds optimal solutions almost always, requiring good run- ning times. However, Greedy + is faster and scales well when changes occur on the input parameters, but provides solutions which are close to the optimum.
Efficient heuristics for data broadcasting on multiple channels
2008
Abstract
The problem of data broadcasting over multi- ple channels consists in partitioning data among channels, depending on data popularities, and then cyclically trans- mitting them over each channel so that the average wait- ing time of the clients is minimized. Such a problem is known to be polynomially time solvable for uniform length data items, while it is computationally intractable for non- uniform length data items. In this paper, two new heuris- tics are proposed which exploit a novel characterization of optimal solutions for the special case of two channels and data items of uniform lengths. Sub-optimal solutions for the most general case of an arbitrary number of channels and data items of non-uniform lengths are provided. The first heuristic, called Greedy +, combines the novel characteri- zation with the known greedy approach, while the second heuristic, called Dlinear, combines the same characteriza- tion with the dynamic programming technique. Such heuris- tics have been tested on benchmarks whose popularities are characterized by Zipf distributions, as well as on a wider set of benchmarks. The experimental tests reveal that Dlinear finds optimal solutions almost always, requiring good run- ning times. However, Greedy + is faster and scales well when changes occur on the input parameters, but provides solutions which are close to the optimum.File | Dimensione | Formato | |
---|---|---|---|
prod_44163-doc_122408.pdf
solo utenti autorizzati
Descrizione: Efficient heuristics for data broadcasting on multiple channels
Tipologia:
Versione Editoriale (PDF)
Dimensione
473.19 kB
Formato
Adobe PDF
|
473.19 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.