透過您的圖書館登入
IP:3.16.147.124
  • 學位論文

可應用於無線感測網路的基因分群演算法在不同情境下與其數據的關係

Clustering by Genetic Algorithm used in wireless sensor network and the data relations in different circumstances

指導教授 : 劉世昌

摘要


無線感測網路為了達到快速傳輸與節能的主要目的,選擇下一個 傳輸節點的位置是十分重要的。通常是以分群方式來節省資料傳輸量, 以群中心做為傳輸的主要節點位置,所以良好的分群結果就可以改善 整體傳輸的效能。 本文以一般常用的距離做為初始分群的依據,結合基因演算法的 部分概念,探討此一演算法在不同情境下其數據的變化,最終結果顯 示了在一定的演化世代範圍內分群數越低、節點數越多,分群結果越 好;節點數越多,收斂速度越快。

並列摘要


For high-speed transmission and energy conservation in wireless sensor network, it is very important to select next node to transfer. Usually, cluster is selected to save energy, and the centers of data clusters are used as the main node of the route to transfer. So that better result in clustering can improve the whole efficiency in transmission. In this study, the common distance initiate clustering, and combines a part of the Genetic Algorithm concept to find the variation of data in different circumstances. Finally, in fixed Generations, the better clustering comes from lower clusters and higher node; the better convergence in higher nodes.

參考文獻


[1]吳宗德,“無線感測網路的節點定位問題之研究”,義守大學資訊
[1]吳宗德,“無線感測網路的節點定位問題之研究”,義守大學資訊工程研究所,pp.11~13,2010。
[2] http://www.myexception.cn/internet/1131568.html。
[3] http://www.360doc.com/content/11/0608/15/7000788_122468694.shtml。
[4] James C. Bezdek1 , Robert Ehrlich2 , William Full3,“FCM: The fuzzy c-means clustering algorithm”, Computers & Geosciences, pp. 191–203,1984。

延伸閱讀