目標物覆蓋(Target Coverage)問題在無線感測網路中一直受到廣泛的討論，本論文考慮監控場景中存在許多我們所感興趣的位置(Points of Interest ( POIs ))，其分散於不連續的區域中，若以固定式感測器佈建整個網路，將佈建數量龐大的感測器，造成佈建成本昂貴，而收集資料時亦造成耗電量不均以及感測器不亦充電的問題。因此，本論文以具移動性之行動資料蒐集裝置(Data Mule)來達成網路的連通及目標物監控或資訊收集等目的。本論文首先提出了Time-Constrained Weighted Targets Patrolling (TCWTP) Algorithm，使分處於各地的Mobile Data Mules以分散式的方式，在POIs之間建置出有效率的巡邏路徑，進行感測及資料蒐集的任務。不同於以往的研究，本論文考慮所有POIs具有不同的權重值，Data Mule對於重要性較高的POIs，將給予其較高的感測與拜訪頻率，以提高其資料更新之速度‧此外，本論文亦針對Mobile Spatial Coverage的議題，考慮監控場景中存在許多不具感測與通訊能力的目標物(POIs)，並解決前述目標物監控所遭遇的硬體成本、目標物監控品質以及Data mules電量等三大挑戰。 This thesis first considers the weighted target patrolling problem which asks a set of mobile data mules (DMs) to efficiently patrol a set of given weighted targets. A patrolling algorithm, called Time-Constrained Weighted Target Patrolling (TCWTP), is presented aiming to construct an efficient patrolling route for a number of given data mules such that targets with higher weights have higher visiting frequency and the overall visiting frequency is stable. Performance study demonstrates that the proposed algorithm outperforms existing approaches in terms of patrolling delay and quality of monitoring.
