Font Size: a A A

Research On Some Key Issues About Energy-Saving In Wireless Sensor Networks

Posted on:2010-12-08Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z MaFull Text:PDF
GTID:1118360278452587Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
In recent years, wireless sensor network technologies have developed rapidly and aroused wide public concern. As one of the core aspects about the performance of wireless sensor networks, energy-saving issue of sensor node has become a hot spot of present research. In this dissertation, four aspects relating to energy-saving in wireless sensor networks are studied, which include node locating, network clustering and routing between clusters, sensing region coverage, and network bottleneck distinguishing. Also corresponding models and algorithms are presented, and analysis and simulation results show their effectiveness. The main innovations in the dissertation are outlined as following:1) A distributed method of WSN node locating, named MDS-MAP(D), is presented. The method based on the theory of Ordinal MDS uses hops as the metric of path length to calculate the shortest path between nodes in local networks derived by local network partition mechanism. Without support of high accuracy ranging technologies and equipments, the method can calculated the relative node location with lower calculating complexity. Furthermore, a method of merging local networks is clearly put forward, which can gets absolute coordinates of all nodes by utilizing the absolute coordinates of limited anchors and gain an increase in locating precision comparing with other similar methods.2) An algorithm for partitioning WSN into non-uniform clusters and building route between clusters, called EER, is introduced. EER controls the node competing radius by a given non-linear function whose behaviors are adjusted by a parameter relating to hops from node to the sink. Under the control of the non-linear function, clusters built in WSN have non-uniform size, i.e., the nearer the sink, the smaller the cluster. The smaller cluster contains less nodes and can save more cluster header energy used for data aggregation for transmitting data between clusters. The non-uniform clustering method can build more clusters near the sink region that helps to balance the energy consumption over different cluster headers.3) A model, which can generate connected coverage with balance energy consumption for sensing region of WSN, is presented. The model based on the Voronoi diagram and the Delaunay triangulation theories includes several links such as hops measurement, nodes layer, network triangulation, sleepable node search and so on. In detail, the model seeks nodes that can sleep together by nodes layer mechanism to obtain the coverage of a sensing region, and ensure the connectivity of nodes at the same time.4) A discriminance method of bottleneck node in WSN, called node bottleneck coefficient, is addressed. The node bottleneck coefficient has two kernel concepts. One is route node probability of nodes and the other is average out-degree of neibour nodes in lower layer. The node bottleneck coefficient characterizes the bottleneck degree of nodes well. And it only needs to know the information of neibours within one hop to calculate the node bottleneck coefficient.
Keywords/Search Tags:wireless sensor network, energy-saving, node locating, node clustering, route, coverage, bottleneck node
PDF Full Text Request
Related items