Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
378992 | Data & Knowledge Engineering | 2010 | 19 Pages |
Abstract
In this paper, we propose new cooperative caching strategies by multiple clients in a push-based broadcast system, which replaces cached items based on benefits of the waiting time. A key idea is that the clients construct a logical peer-to-peer network and each of them determines the replacement of its own cache by taking into account access probabilities to data items from neighboring clients and the broadcast periods of data items or the times remaining until these items are broadcasted next. We confirm that the proposed strategies reduce the average response time by simulation experiments.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Takahiro Hara, Kazuhiko Maeda, Yoshimasa Ishi, Wataru Uchida, Shojiro Nishio,