無線感測網路的移動錨節點定位問題之研究

The Study of Mobile Anchor-Based Localization in Wireless Sensor Networks

10.6343/ISU.2014.00209

王聖博

移動錨節點 ； 高斯帕曲線 ； 節點定位法 ； 無線感測網路 ； 路徑規劃 ； Mobile Anchor ； Gosper Curve ； Localization Problems ； Wireless Sensor Networks ； Path Planning Algorithm

義守大學資訊工程學系學位論文

2014年

碩士

陳啟彰

繁體中文

在本論文中，我們提出一個可以擴展的路徑規劃演算法來作為錨節點(Mobile Anchor)的移動軌跡，採用空間填充曲線─高斯帕曲線(Gosper Curve)，讓自走車或機器人載著裝配GPS的錨節點沿著預設軌跡移動，並且在其六邊形網格中心點廣播位置訊息，我們的方法可以保證均勻遍歷整個感測區域。另外我們提出一種高斯帕島 (Gosper Islands)的座標系統，錨節點可以確定下一個廣播座標，實現抗障礙物的機制。如果使用相同錨節點感測半徑，高斯帕曲線作為錨節點移動路徑的方法比其他方法使用更少的廣播次數與移動路徑長度，更節省能源消耗。對於無線感測網路的錨節點定位問題，我們使用三點定位法搭配最小平方法來提升定位精準度，六邊形網格的排列方法可以有效的避免錨節點共線性問題。本論文的實驗以實際測量接收訊號強度值(RSSI)估測距離，並且以平均距離加上一到三倍的測量標準差進行模擬，當使用最大的RSSI值標準差時，定位誤差約為傳輸距離的百分之四點五，明顯較其他方法為佳。

In this thesis, we propose a scalable path planning algorithm as the trajectory for the mobile anchor in a wireless sensor network. We use Gosper curve, a space filling curve, as a trajectory for the robocar, equipped with GPS, to broadcast location information while travelling to the center of each hexagon along the path. Our approach can ensure that the mobile anchor traverses the entire sensing area uniformly. Meanwhile, we introduce a coordinate system of Gosper Islands for deciding the broadcast position of the mobile anchor, especially when the path has obstacles. If we use the same size of circumscribed circle of radius r for each grid (triangle, square, or hexagon), the anchor node moving path with Gosper curve use less broadcast time, length of moving path, and energy consumption than other methods. For anchor node localization problem in wireless sensor networks, we use trilateration with the least squares method to estimate the positions of unknown nodes. The method of the hexagonal grid arrangement can effectively avoid the anchor node co-linearity. In this thesis, we use the actual measured value of RSSI (Received Signal Strength Indicator) to estimate the distance between anchor node and each unknown node, and the simulations are with one to three times standard deviation of distance errors. Even when we use the maximum value of the standard deviation of RSSI for the distance errors, the location error is only about 4.5% of the transmission distance. It is better than most of the contemporary localization methods.

1. 謝政達(2016)。碎形幾何應用於無線感測網路之研究。義守大學資訊工程學系學位論文。2016。1-103。
2. 陳寶珠(2016)。無線感測網路的目標追蹤之研究。義守大學資訊工程學系學位論文。2016。1-64。