On Sequential k-coverage Verification

• Main Authors: Juang, Kun-Han, 莊坤翰
• Other Authors: 陳良弼
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/94120895671332920176

碩士 === 國立清華大學 === 資訊工程學系 === 100 === One fundamental problem in sensor networks is the coverage problem, which reflects the degree of a sensor network being monitored or tracked by sensors. We say a point is k-coverage which means it is monitored by at least k sensors. The previous works on k-coverage verification mainly concentrate on two types: 1) checking a set of discrete target points and 2) checking the whole target plane. Consider a scenario as follows. An important road segment is monitored by sensors and we need to verify whether the whole segment is covered by at least k sensors to guarantee the quality of service. To the best of our knowledge, there are no existing methods considering the coverage verification regarding a line segment. In this thesis, we make the first attempt to issue the problem of k-coverage verification regarding a line segment. To deal with the problem, we propose three methods based on the R-tree index. The Basic method is the most primitive method which calculates all the intersection points of the monitoring regions of the sensors to check whether each sub-segment is k-covered. The Projection method uses the lower bound of the cover range of sensors, and reduces computation cost because of the coordinate value stored in advance. As for Tessellation method, we use the “pruning window” such that all the points in this range can cover a specific interval of query line. This method only considers a few necessary sensors to verify the coverage of the query line. A series of experiments are performed for evaluating the proposed methods .The experiment results demonstrate that Tessellation has the best performance.