| dc.description.abstract | 無線動態隨意網絡(Mobile Ad hoc networks) 是一群節點藉由網絡傳輸連結而組成, 其中有一部份節點, 我們稱之為關節點(Articulation points), 也就是說, 在網絡中將移開關節點, 網絡將會分裂成一個以上不同的子網絡, 導致傳遞的資訊無法互相溝通. 相關文獻提出在靜態的網路(Static Networks) 之下偵測關節點, 然而在動態的網路 (Dynamic Networks)之下快速和準確的偵測關節點是一個挑戰. 這篇博士論文動機緣起於, 報章新聞報導一群人出遊, 常有走失現象, 例如: 老師帶小朋友出遊, 老人出遊, 甚至是車隊出遊, 這樣的活動由於人數多, 在管理和監控上不易, 基於此首先我們提出了區域深先搜尋法(Localized Depth First Search)來解決在動態網路下偵測關節點這個問題. 接下來考慮到省電(Power Saving)的議題, 在動態網路之下, 無線網路裝置的電量是有限的, 如何在不影響關節點偵測之下, 達到最省電的機制, 我們提出了省電搜尋法(Power Saving-LDFS), 實驗結果表示, 我們的方法在訊息負載, 偵測速度 ,電量節省上皆有較優秀的表現. | zh_TW |
| dc.description.abstract | The articulation points in the induced graph of a network are known to be critical in terms of maintaining the connectivity of the network as most of the routing paths between nodes in the network have to pass through them. To detect articulation points, the previous works assume the setting of static networks. However, in case when a group of people move together, their mobile devices forms a highly dynamic mobile ad hoc network and the behavior of nodes in such a network can be a challenge when it comes to detecting articulation points. In this paper, we propose a distributed algorithm, namely Localized Depth-First-Search (LDFS), to identify articulation points efficiently in mobile ad hoc networks. The algorithm sets multiple initiators to send probing packets and a bound for the hop count of the probing packets in order to improve the accuracy rate and message overhead at the same time. The simulation results show that our LDFS algorithm outperforms the other well-known protocols, which suggest that LDFS is a better solution for articulation point detection to prevent possible losses in a group of moving nodes. In second part, we proposed a Power-Saving LDFS protocol to save power of node and still maintain the connectivity of network. The performance of energy consumption is better than original case. | en_US |