「無線感測網路」被廣泛應用於環境監控、國土防衛等多重領域，多數應用均需要感測節點瞭解其本身之位置資訊，因感測節點通常係隨機且大量布放，故在既有條件下提升定位效能已成為重要研究課題。 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，然無線干涉定位方法亦有運算負荷大、可能產生模糊結果等缺點。先前之無線干涉定位相關演算法對初始階段網路結構之假設要求過於強烈，且常忽略運算複雜度過大之問題，在本論文中我們以「無線干涉定位系統」為研究對象，先就其它相關演算法特性進行比較，再提出更具網路適應性且不增加運算負擔之演算法，並以模擬方式驗證其效能。; Wireless sensor network（WSN）has been widely applied to multiple fields such as environment monitoring, homeland security etc. Most of these applications need sensor nodes to be aware of their location informations. Since the sensor nodes in WSN are usually deployed in large numbers, and often randomly, the way to find locations of these sensor nodes precisely and efficiently thus becomes an important issue. There are some properties of radio interferometric positioning system（RIPS）such as high accuracy, no extra hardware cost and applicable in both indoor and outdoor circumstances; therefore, it is very suitable for WSN localization. However, heavy computing complexity and localization ambiguity under specific conditions are with coming problems. The former studies of RIPS algorithms usually neglect the calculation load of localization process and require strict presumptions of network configuration. In this paper we compare the RIPS algorithms with these concerning characteristics and propose an improved algorithm more adaptive to network topology without increasing computing complexity. The simulation results of algorithm efficiency comparison are also demonstrated.
