無線感測網路中，定位技術是極其重要的一個角色，在其應用領域內，事件發生的位置是感測節點監測訊息中的重要訊息，因位沒有節點位置資訊的數據是毫無意義的。 無線干涉定位技術因具有準確性高、無需額外增加設備成本、室內外環境皆可運用等特性，非常適合做為無線感測網路之定位方法，然而無線干涉定位方法在測量未知節點時還是有可能會產生模糊結果。在本文中，我們提出了一種利用多組RIPS測量所得的結果，來確定未知節點的真實位置的定位演算法。並且此演算法是以分散式系統架構設計的，並且在當RIPS測量發生模糊結果時才會需要透過額外的RIPS測量用以獲得更多結果來找到節點真正的位置。而在我們的實驗中證明了此演算法在適當的節點密度或足夠數量的錨節點的情境下可以成功辨識90%以上的未知節點。 ;Localization is one of the fundamental services in wireless sensor networks. Many applications, such as environmental monitoring and tra c control, rely on the avail- ability of location information for wireless sensors. Among all types of localization techniques, the Radio Interferometric Positioning System (RIPS) is promising due to its high accuracy and low hardware requirements. However, when RIPS is ap- plied to a large-scale wireless sensor network, at times the RIPS measurements may lead to ambiguous positioning for a free node, i.e., there may be more than one pos- sible position for a free node. If the free node is assigned an incorrect location, the error will be propagated to other free nodes, leading to increasing localization errors in the network. In this thesis, a localization algorithm is proposed which utilizes redundant location information from multiple cases to identify the true location of free nodes. The proposed algorithm is fully distributed, and the extra overhead is paid only when the ambiguity issue occurs. The simulation results demonstrate that the proposed algorithm is able to localize more than 90% of the free nodes with adequate node density or enough number of anchors.
