無線感測器網路 (Wireless Sensor Network, WSN) 領域近年來受到越來越多的關注，由於無線感測網路擁有低成本、低頻寬、低能源耗損以及防碰撞機制，促進了許多新的應用。位置定位是其中非常重要的一環，如何提供準確的位置資訊，是近年來很熱門的研究主題。 本篇研究是在無線感測器網路中，利用接收訊號強度 (Received Signal Strength, RSS) 技術來對目標物進行追蹤，並使用擴展型卡爾曼濾波器 (Extended Kalman Filter, EKF) 來過濾 RSS 的變化，改善移動中目標的位置估計。除此之外，傳輸環境中可能存在數個目標物或雜波干擾的問題，為了降低干擾所造成的影響，我們使用機率數據關連濾波器 (Probabilistic Data Association Filter, PDAF)、機率假設密度濾波器 (Probability Hypothesis Density Filter, PHDF)來改善此問題，各個感測器將所接收到的訊息彙整至數據融合中心 (FusionCenter, FC)，再由數據融合中心計算出較佳目標物路徑軌跡。;Wireless sensor network (WSN) is an active research area that has attracted much attention in recent years. Since the sensors used in a WSN have the properties of low cost, low bandwidth, low energy consumption, and anti-collision mechanism, WSN has been found in many applications. How to know the accurate positions of mobile terminals in a WSN is an important issue. This thesis studies an Received Signal Strength (RSS) technique to track mobile targets in a WSN and employs the Extended Kalman Filter (EKF) for position estimation of moving targets. In addition, there are usually multiple targets and clutter interferences in the tracking environment. To reduce the impact of interferences, we consider the Probabilistic Data Association Filter (PDAF) and Probability Hypothesis Density Filter (PHDF) to improve the tracking problem. Then, a data fusion center (FC) calculates target tracks with the weighted least squares method from the messages provided by multiple sensors.
