Please use this identifier to cite or link to this item:
|Title:||Batch scheduling for demand-driven servers in wireless environments|
|Authors:||Tan, K.-L. |
|Citation:||Tan, K.-L.,Ooi, B.C. (1998-08). Batch scheduling for demand-driven servers in wireless environments. Information Sciences 109 (1-4) : 281-298. ScholarBank@NUS Repository.|
|Abstract:||In a demand-driven information/database server, the server transmits data to clients that explicitly request for them. Because of the low bandwidth of wireless networks, the number of clients that can be effectively supported is often limited. Caching has been shown to be effective in reducing the contention on the narrow bandwidth of the wireless channel. However, it cannot handle contention due to multiple clients requesting the same data objects. In the latter case, the server should "intelligently" transmit only one copy of the data objects to satisfy multiple requests that arrive within a short duration of time. In this paper, we examine the effectiveness of batching in increasing the number of clients that wireless networks can support. Under a batching scheme, data objects may not be transmitted immediately even if the channel is available; instead, they may be delayed so that multiple queries for the same data objects can be served by the same single download. In static batching, the delay is predetermined and fixed. On the other hand, under an adaptive batching scheme, the delay varies depending on the heuristics employed in selecting the data objects to be transmitted. Additionally, if clients' connection time are short and predictable, we can further enhance the adaptive schemes to discard data objects to better utilize the channel. A simulation study is conducted, and we report on the results in this paper, © 1998 Elsevier Science Inc. All rights reserved.|
|Source Title:||Information Sciences|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Nov 9, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.