臺灣大學: 資訊網路與多媒體研究所黃寶儀黃俊維Huang, Chun-WeiChun-WeiHuang2013-03-222018-07-052013-03-222018-07-052010http://ntur.lib.ntu.edu.tw//handle/246246/251329YushanNet是建設在玉山國家公園的耐延遲無線感測網路，用以追蹤登山客軌跡，具有山難救助與遊客遊憩功能。其設計有CSMA與TDMA兩種不同的頻道存取協定。本論文是針對CSMA所做的研究。 有效率地設定CSMA參數是一個重要的議題，為此我們利用分析登山客移動的特性達到參數設定的方式可有效縮減需要模擬的參數設定範圍。我們採用GPS logger記錄玉山登山客的GPS track data來作為資料分析的依據。Trace分析結果發現，登山客多數相遇發生在西峰涼亭附近和接近中午左右。相遇時間都很短暫且相遇的頻率約半秒鐘到五分鐘。另外，相遇的neighbors會隨不同的時間而改變，網路節點變動情形也很頻繁。 最後，模擬結果顯示利用Mobility分析的參數設定可有效降低模擬參數範圍。雖然與Common sense推測方法的模擬結果很接近，但透過分析找到一些重要的mobility特性卻可提供未來在hiker-aware DTN設計的最佳化。YushanNet is for hiker tracking in the Yushan national park. YushanNet which is a delay-tolerant wireless sensor network will apply for mountain rescuing and tourist information. There are contention-free (TDMA) and contention-based (CSMA) MACs for YushanNet. Our research addresses the CSMA. The effectiveness of parameter settings is an important issue. Our methodology is using hiker mobility analysis to effectively minimize the range of parameter settings. We use GPS loggers to collect hikers’ track data for mobility analysis. The results of trace analysis showed that the most of encounters took place at around noon and near the West Peak Pavilion. The encounter duration was short, and the hiker encountered another hiker at frequency of 30~300 seconds. In addition, the encounter neighbors of nodes varied from time to time, and the member nodes of network were variable. Simulated results show that it is feasible to minimize the range of parameter settings via mobility analysis. Although the results between our methodology and common sense inference are close, our methodology can find some mobility characteristics for optimizing hiker-aware CSMA in future.1831915 bytesapplication/pdfen-US無線感測網路耐延遲網路參數設定移動性分析媒體存取控制Wireless Sensor NetworkDelay Tolerant NetworkParameter SettingsMobility AnalysisMedium Access Controlthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/251329/1/ntu-99-R97944026-1.pdf