延長網路存活時間之適應性錨節點睡眠排程

Abstract

定位技術已在無線感測器網路(WSNs)中被廣泛應用於找出節點之未知位置。一般在執行定位技術的過程中，會大量佈置錨節點(位置已知之節點)來協助推算其他節點之位置；然而，同時啟動所有的錨節點並不能明顯地增加定位準確率，反而會增加額外的能量成本與頻寬成本。在這樣的情況下，通常只需要同時啟動一部分的錨節點就能達到準確率的要求，因此為了減少成本並延長網路存活時間，本論文提出了可使錨節點自己安排工作周期的Adaptive Beacon Duty Scheduling (ABDS)演算法。ABDS會在線上即時量測錨節點位置之效益(在此位置啟動錨節點後可能帶給定位準確率多少助益)，並且根據此量測結果挑出對定位準確率最有效益的那些錨節點以啟動之，以將同時啟動的錨節點數量最小化。由於在前人的相關研究中，並未實際去量測不同錨節點位置的不同效益，因此ABDS可更佳地適應充滿無法預測之雜訊的真實環境。此外，為了在ABDS中精確地量測錨節點位置之效益，我們觀察到錨節點對其覆蓋範圍產生的定位效能改善量事實上是非均勻分布的，並提出了尚未被討論過的Distribution-Adapted Grid (DAG)量測法以適應此現象。與前人的方法相比，使用了DAG量測法的ABDS可以減少10%的錨節點使用量，並且延長54%的存活時間。

Within typical localization processes in wireless sensor networks (WSNs), beacon nodes which know their locations will broadcast information for localizing an unknown location. Although beacon nodes are massively deployed in the terrain, only a fraction of the beacon nodes are required to be active for satisfying accuracy requirement. Too many active beacon nodes may bring the system with little improvement on localization accuracy but waste of both costs of energy and bandwidth. To reduce the costs and prolong the system lifetime, we propose the Adaptive Beacon Duty Scheduling (ABDS) algorithm that can self-configure beacon duty. ABDS can turn on only the minimum set of beacon nodes in a same time according to the online-measured effectiveness of beacon locations (the effect of activating a beacon node at the location for improving localization performance), which is not considered in previous methods. Moreover, to precisely measure the effectiveness of beacon locations in ABDS, we need to realize the fact that a beacon node actually contributes non-uniformly distributed impact within its coverage. This Distribution-Adapted Grid (DAG) measurement that can adapt the non-uniformly distributed impact was not discussed in previous methods. Compared to the previous methods, ABDS with the usage of DAG measurement can reduce 10% beacon usage and provide 54% longer lifetime.