The barrier coverage mechanisms for improving the surveillance quality in wireless visual sensor networks

機構典藏 > College of Engineering > Graduate Institute & Department of Computer Science and Information Engineering > Thesis > Item 987654321/114673

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/114673

Title:	The barrier coverage mechanisms for improving the surveillance quality in wireless visual sensor networks
Other Titles:	在無線視覺感測網路中具提升監控效能之邊界覆蓋技術
Authors:	蕭志堯;Hsiao, Chih-Yao
Contributors:	淡江大學資訊工程學系博士班張志勇;Chang, Chih-Yung
Keywords:	k-Barrier Coverage;k-邊界覆蓋;QoS;Surveillance;Wireless Sensor Networks (WSNs);Wireless Visual Sensor Networks (WVSNs);無線視覺感測網路;無線感測網路;監控品質
Date:	2017
Issue Date:	2018-08-03 15:00:21 (UTC+8)
Abstract:	無線視覺感測器網路（WVSNs）由許多視覺感測器組成。市面上有許多類型的視覺感測器已被廣泛應用在物聯網應用中。這些類型的視覺感測器包括：紅外感測器，微波感測器，超聲波感測器，照相機和雷達。與全向感測器相比，視覺感測器擁有更好的性能，因為它可以進一步回傳入侵者的方向。然而，大多現有的邊界覆蓋機制皆只考量在全向性的無線感測器網路(WSNs)中，因此不能有效地應用在無線視覺感測器網路，而原因在於視覺感測器的感測區域為扇形感測區域。有鑑於此，本論文研究在無線視覺感測器網路中，提出具監控效能提升之邊界覆蓋技術。該技術有三種演算法，分別是：基礎演算法(BA)，回溯演算法(BTA)和分支演算法(BRA)，旨在建構出數量最大且彼此不相交集之防禦屏障，其中，每條建構出的防禦屏障亦能支持k-邊界覆蓋監控品質之需求，如此將能夠確保入侵者的任意入侵路線，皆能夠被視覺感測器偵測至少k次，以提升監控可靠性。針對本論文所提出之邊界覆蓋技術，我們除了以理論的角度分析其性能外，亦透過實驗模擬與其他現有相關論文進行多維度的比較。而分析與實驗結果皆顯示本論文所提出之邊界覆蓋技術實現了與最優解相似的性能，且具有較少的控制封包數量。除此之外，與現有論文相比，本論文所提出之分支演算法不僅在控制封包數量上具有絕對優勢外，在防禦屏障建構的數量上更具有較佳的性能。 A wireless visual sensor network (WVSN) consists of a number of visual sensors. Many types of visual sensors have been widely applied in constructing a wireless sensor network for the IoT application. These types of sensors include infrared sensor, microwave sensor, ultrasonic sensor, camera, and radar. Compared with the omni-directional visual sensor, the visual sensor can achieve better performance because it can further report the direction of the intruder. Unfortunately, most existing barrier-coverage mechanisms considered the omni-directional sensor networks. They cannot be efficiently applied to WVSNs because the sensing area of each visual sensor is fan-shaped. This dissertation investigates the surveillance service problem which supports surveillance quality of k-barrier coverage in WVSNs. Three algorithms, called BA, BTA and BRA, are proposed aiming at finding a maximal number of different defense barriers, each of which supports k-coverage. As a result, the intruders intending to cross the monitoring area can be detected by at least k visual sensors. Performance analyses of the proposed algorithms are conducted in chapter 3.3.3 to verify the performance improvement from the theoretic point of view. In performance study, the proposed algorithms are compared with other existing works. Experimental study shows that the proposed k-barrier coverage algorithm achieves similar performance to the optimal solution but has fewer control packets. Furthermore, the proposed BRA achieves better performance in terms of the numbers of control packets and constructed defense barriers, as compared with the existing works.
Appears in Collections:	[Graduate Institute & Department of Computer Science and Information Engineering] Thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	149	View/Open

Loading...