Research On Node Localization Of Wireless Sensor Network

Nowadays, with the development of science and technology, wireless sensornetwork application has gradually infiltrated into almost all fields of informationsociety, and the node localization is the focus of attention. Positioning can be dividedinto two categories: location rang base and range-free algorithm. The later algorithmdoes not require additional equipment, and has advantages in energy consumption,cost, and is more suitable for fileds of low power consumption and low cost. So ourresearch is based on range-free algorithm.Range-free algorithm is usually called as distance-independent algorithm.Current available distance-independent algorithms are divided into two categories.The first kind of algorithm work as follows: firstly, it estimates the distance betweenunknown nodes and beacon node, then locates the unknown node using the trilateralmeasurement method or the maximum likelihood estimation method. The other kindof algorithm work as follows: the area of unknown node is determined firstly throughthe location of neighbor node and beacon node, then we take the centroid of the areaas the coordinates of the unknown node. Though distance-independent algorithm haslow accuracy, it can meet the requirements of most applications. Therefore, this thesisprovides two algorithms based on range free localization.DV-Hop is a typical range-free localization algorithm, on which many otheralgorithms are based. Given the limitations of DV-Hop algorithm, the followingoptimizations are studied in this thesis.1) We proposed a localization algorithm based on average distance. The averragehop distance (AHD) of each anchor node is caculated by the distances of one-hopneighbor anchor nodes and the boundary anchor nodes, then the anchor nodesbroadcast their AHD to the whole network. Considering the particularity of networkdistribution, after receiving the packages including AHD from all anchor nodes，theunknown nodes set a weight factor for each anchor node according to the hop-count tothe anchor node. Experiments show that twice improvements to AHD raise thepositioning accuracy.2) We utililize power-control to realize node localization. The anchor nodes broadcast their position information under different transmission power, and theunknown nodes estimate the distance to the anchor nodes according to the receivedpower. The unknown nodes can calculate their coordination when obtaining enoughdistance information. Further more, in order to achieve higher accuracy, RSSI is usedto correct the estimated distance.Given enough transimission power, three anchornodes can be used to achieve node-localization in wireless sensor network area.Considering the low accuracy of the traditional centroid algorithm, this paperputs forward an improved weighted centroid algorithm, then self-corrected. We takeany four beacon nodes to improve the node localization accuracy, retaining thesimplicity of algorithm and saving nodes energy. These four beacon nodes compose aquadrangle, which is then divided into two triangles to seek centroids and get theiraverage values. Thereafter, we measure each beacon node and empower the values,then use the centroid localization algorithm again to locate object nodes. Thesimulations show that the algorithm effectively improves the positioning accuracy ofthe unknown nodes.