有邊界監測環境下異質方向性行動感測器之加強型佈署演算法設計與實作

Abstract

在無線感測網路裡，良好的感測器佈署機制能使整體網路獲得較為完整的監控能力。相較於全向性感測器，方向性感測器的感測範圍跟感測半徑(距離)、感測方位(視角)、感測角度與精準度有關。由於方向性感測器中可能具有各種不同的感測半徑與角度，我們稱之為異質方向性感測器。在本篇論文中，我們針對一個具有邊界條件的監控環境下，使用具有移動和旋轉特性的異質方向性感測器進行自我佈署，並搭配兩種加強型佈署演算法: EDA-I 和EDA-II，在真實的環境佈署異質方向性感測器，提高整體環境的覆蓋率。EDA-I利用虛擬力(virtual force)和虛擬邊界力矩(virtual boundary torques)的概念分別來完成感測器的移動與旋轉；而EDA-II 則是使用Voronoi圖引導感測器移動，並使用邊界力矩旋轉感測器至適當的角度。最後，為了實際比較並驗證這兩種演算法的實用和正確性，我們在使用一台電腦作為中控端伺服器的集中式運算模式環境下，裝載所提出的演算法於Arduino自走車的平台上來進行測試。而經由實測的結果，我們可以看出此兩種演算法確實可以被有效地應用在真實的監控環境。

In wireless sensor network, it is important to improve coverage via good deployment of sensors. Different from omnidirectional sensors, the sensing range of directional sensors depends on the sensing radius, spread angle, orientation and accuracy. Since directional sensors have different sensing parameters, which leads to different sensing ranges, we adopt heterogeneous directional sensors model in this research. In this thesis, to monitor a bounded area, we use heterogeneous directional sensors with moving and rotating abilities to execute two self-deployment algorithms: EDA-I and EDA-II, to improve sensing coverage in a real-life prototype. EDA-I uses virtual forces and virtual boundary torques to control the movement and rotation of the sensors, while EDA-II controls sensors based on Voronoi diagram and virtual boundary torques. Finally, in order to verify the performance of these two algorithms, we use an off-the-shelf computer as the controlling server with two deployment algorithms and Arduino robots in a centralized architecture. The experimental results demonstrate that our deployment algorithms can be applied in a real environment.