無線感測網路指向天線定位機制; Localization with Directional Antennas in Wireless Sensor Networks

NCU Institutional Repository > 資訊電機學院 > 資訊工程研究所 > 博碩士論文 > Item 987654321/63527

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/63527

题名:	無線感測網路指向天線定位機制;Localization with Directional Antennas in Wireless Sensor Networks
作者:	林志明;Lin,Chih-ming
贡献者:	資訊工程學系
关键词:	指向天線;收訊階度;收訊角度;指紋;定位;迴歸分析;Directional antenna;Received Signal Strength Indicator (RSSI);Angle of Arrival (AoA);Fingerprinting;Localization;Regression Analysi
日期:	2014-01-20
上传时间:	2014-04-02 15:46:58 (UTC+8)
出版者:	國立中央大學
摘要:	本論文運用指向性天線的特性，設計並實作旋轉指向天線定位(Rotatable Antenna Localization, RAL)、收訊階度差值收訊角度定位(AoA Localization with RSSI Difference, ALRD)與十字天線陣列指紋定位(Fingerprinting Localization with Cruciate Directional Antennas, FLCDA)等三種無線感測網路(Wireless Sensor Network, WSN)定位機制。RAL定位機制以兩個裝配可旋轉指向天線的信標節點(beacon node)或錨節點(anchor node)協助目標節點(target node)(即位置未知的感測器節點)完成定位，信標節點將指向天線旋轉到不同角度，並發送包含其位置與天線角度的信標信號；而目標節點透過觀察所接收信標信號強度的變化，估算其與信標節點的角度，當獲得兩個不同位置信標節點的角度與位置時，目標節點就能自行估算其所在位置。此外，我們提出田字基礎(grid-based)與向量基礎(vector-based)兩種改進方法，藉由增加信標節點以降低定位誤差(localization error)。 ALRD定位機制以兩支相互垂直指向天線的收訊階度(Received Signal Strength Indicator, RSSI)差值估算收訊角度(Angle of Arrival, AoA)，進而完成感測器的定位。本機制區分測量與定位兩個階段，測量階段為在系統部署前實際量測收訊角度對指向天線收訊階度的影響，並分別以二次與線性迴歸分析取得在不同距離下，收訊角度與收訊階度的關係函數，以及相互垂直天線的收訊階度差值與收訊角度的關係函數，並將所得的關係函數儲存於目標節點。在定位階段，我們在環境中部署二個已知位置的信標節點，每個信標節點上裝配著兩支同型且方向相互垂直的指向天線。目標節點在接收到同一個信標節點的二支天線所發出的訊號後，先利用收訊角度與收訊階度的關係函數概估與信標節點的距離，再運用該距離所相對應的收訊階度差值與收訊角度的關係函數，計算與信標節點的角度。在取得對兩組已知位置信標節點的角度後，目標節點即可計算出自己的位置。我們更進一步提出最小直徑涵蓋最多估算點(Maximum-Point-Minimum-Diameter, MPMD)與最小面積涵蓋最多估算點(Maximum-Point-Minimum-Rectangle, MPMR)兩種方法，透過選取多次定位結果中密度最高的估算點，再以這些估算點的距心(centroid)做為最後的目標節點位置以降低定位誤差。 FLCDA的錨節點裝配四支方向相互垂直的指向天線，並以四支天線同步接收由目標節點所發送的信號，以協助目標節點完成定位。FLCDA定位區分測量與定位兩個階段，在測量階段，錨節點旋轉方向天線度(=0,1,..,359)，接收來自d公尺(d=0.5, 1,…,6)之外感測器節點的信號。一對(, d)配對被視為一個參考位置(reference position)。錨節點記錄四支天線所接收信號的收訊階度作為不同參考位置的指紋(fingerprint)，並將所有參考位置與其指紋存放在錨節點的資料庫中。在定位階段，錨節點接收目標節點的信號，再與資料庫中指紋進行比對，以最吻合指紋的位置做為目標節點的估算位置。由於十字天線陣列指紋定位機制需要較長的時間蒐集指紋與定位，我們設計兩種改進的方法分別用於縮短指紋蒐集時間與定位時間。我們在室內體育館實作三種定位機制的實驗，RAL與ALRD定位機制的實驗範圍為10米見方的方型區域，FLCDA定位機制則是直徑12公尺的圓形區域。實驗結果顯示三種機制的平均定位誤差分別為76、124與24.4公分，RAL與ALRD定位機制的改進方法可分別降低約10%與29%的誤差。因為RAL與ALRD定位機制可以讓目標節點自行定位，因此它們可用於需隱藏目標節點位置資訊的應用場景；而RAL定位機制因為不需要收集訊號資料的測量階段，因此其定位準確性較不受溫、濕度等環境因素改變所影響，也不會因天線發送或接收增益的不同有大幅度的改變。ALRD與FLCDA定位機制僅需少量信標信號即可完成定位，可以減少電源的損耗，也可以用來追蹤感測器節點的移動。此外，在FLCDA定位機制中，單一錨節點即可獨立完成目標節點的定位，這可方便我們透過部署大量的錨節點，將定位機制應用於較大範圍的定位環境。; In this dissertation, three RSSI-based localization schemes, Rotatable Antenna Localization (RAL), AoA Localization with RSSI Difference (ALRD) and Fingerprinting Localization with Cruciate Directional Antennas (FLCDA), are proposed for localizing sensor nodes in wireless sensor networks (WSNs). In RAL, we employ two beacon nodes whose positions are known and which are equipped with directional antennas rotating regularly to localize target nodes (i.e., the sensor nodes whose positions are unknown). The beacon nodes periodically send beacon signals containing their positions and the orientations of the antennas. By observing the variation of the received signal strength indicator (RSSI) values of the beacon signals, a target node can estimate the orientation relative to the beacon nodes. With the estimated orientations and exact positions of two distinct beacon nodes, the target node can calculate its own location or position. Moreover, we try to propose reduce localization errors by installing more than two beacon nodes and by using grid- and vector-based approximation methods. ALRD estimates the Angle of Arrival (AoA) for localization by comparing the RSSI values of beacon signals received from two perpendicularly oriented directional antennas installed at the same place. ALRD consists of two phases: the RSSI gathering phase and the localizing phase. In the RSSI gathering phase, we measure the RSSI values of signals received from a directional antenna at different distances and angles. For a fixed distance d, we perform regression analysis on the measured RSSI values to obtain two approximation functions: a quadratic function Rd = f(θ) and a linear functions Dd = g(θ), where θ is AoA, Rd is RSSI, and Dd is the RSSI difference of two signals received from two perpendicular directional antennas at the same position. These approximation functions are then loaded into the limited storage of target nodes to calculate AoA values. In the localizing phase, a target node estimates the distance and AoA to a beacon node by the approximation functions and the RSSI values received from two antennas of the beacon node. With the estimated AoA values of two distinct beacon nodes, the target node can then calculate its position. We further propose two methods, named maximum-point minimum-diameter (MPMD) and maximum-point minimum-rectangle (MPMR), to reduce localization errors by gathering more beacon signals within 1 s for finding the set of estimated locations of the maximum density. The centroid of the estimated locations in the set is calculated to be the final location of the target node. In FLCDA, an anchor node which is equipped with four directional antennas arranged as a cross or cruciate shape (i.e., the orientations of the antennas are perpendicular to the adjacent ones) can alone localize target nodes. FLCDA consists of two phases: the RSSI gathering and the localizing phases. In the RSSI gathering phase, the anchor node's antennas are rotated by  degrees (say, =0,1,..,359) to receive signals sent by a sensor node which is d meters (say, d=0.5, 1,…,6) away from the anchor node. The pair (, d) is regarded as a reference position, and the RSSI values of signals received by the four antennas are regarded as the fingerprint of the position. All reference positions and their associated fingerprints are stored in the anchor node before deployment. In the localizing phase, a target sensor node sends a signal to request the anchor nodes to help with localization. On receiving the request signal from the four directional antennas, an anchor node finds out the fingerprint most matched with the received RSSI values. The position associated with the most matched fingerprint is assumed to be the target node's position. Because FLCDA needs lots of time for collecting fingerprints and localizing, we also design and implement some techniques to accelerate FLCDA without affecting the localization accuracy too much. We implement the above schemes and apply them to a WSN in an indoor gym for conducting experiments. RAL and ALRD are installed in a 1010 meters area while FLCDA is installed in a 12-meter-diameter circle area. Our experimental results show that the average localization errors of RAL, ALRD and FLCDA are 76, 124 and 24.4 centimeters, respectively. The proposed improvement methods of RAL and ALRD can reduce the average localization errors by factors of about 10% and 29%, respectively. Since RAL and ALRD allow target nodes to calculate their positions on their own, RAL and ALRD can be applied to a WSN that the position information of target nodes must be kept secret. RAL has no RSSI gathering phase and thus the variations of temperature, humidity and the gains of the transmitters' or receivers' antennas has little influence on localization errors. ALRD and FLCDA can be applied to a WSN in which the sensor node power consumption is critical because they need only few signals for localization. They are also suitable for tracing mobile sensor nodes because they can fast localize sensor nodes. Moreover, FLCDA can be easily applied to large areas by deploying numerous anchor nodes because it needs only one anchor node to localize target nodes.
显示于类别:	[資訊工程研究所] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	449	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....