Research On Key Technology Of Localization For Wireless Sensor Network

Wireless Sensor Network (WSN) refers to the distributed self-organizing network comprised by a large number of sensor nodes with the sensing, computing and wireless communication capabilities, which has been widely used in environmental monitoring, security monitoring, health care and other fields. Along with the further development of WSN, its characteristics and advantages have become increasingly significant, it has been more and more widely applied, and as one of the key technologies of location estimation has also attracted great concern and attention. The node location information is usually the precondition to realize numerous applications of WSN, and its accuracy is one of key performance indicators of WSN. Therefore, the node location technology has an important role in WSN.The basic features of WSN are energy-efficiency and self-organizing. It obviously won’t work by installing GPS on all nodes to achieve locating, and in addition, GPS only applies to the outdoor condition without any covers. Therefore, we can only install GPS on some of the nodes, and the location of other nodes will be estimated through certain algorithm. After years of development, various approaches to locate nodes have been proposed, and great effects have been achieved under certain conditions. However, like many key technologies, there are many technological problems left to be solved in the node localization technology. This is due to the bottleneck existed in the node localization technology:1) The environment of monitoring area is complicated, and it is affected by various factors such as the environment, barriers and attacks, which has caused the instability of measuring accuracy.2) Because the node deployment is usually random, when the beacon nodes used for location estimation are close to be in the same line or same plane, it will cause the problem of incalculable unknown nodes.3) Because the monitoring area is relatively wide, the node communication radius is limited, and the beacon nodes are randomly deployed, it makes it difficult for some unknown nodes to gain enough beacon nodes to estimate location. As a result, the coverage rate of localization algorithm is low.4) Traditional localization mechanism always carries out location estimation by using the trilateral or multilateral method through adjacent beacon nodes. The measuring error will increase when the beacon nodes are far away from the unknown node, which will ultimately cause higher estimation error. Based on those factors mentioned above, this dissertation has conducted in-depth research in accordance with the problems mentioned above. The main research contents and innovative points include the following:1. A weighted median-based ranging localization algorithm is proposed based on analysis of the characteristics of ranging noise. Each measured data is fully used by the algorithm, and based on the median, different measured data are assigned with corresponding weights to reduce the impact outliers and smooth random error in measuring data.2. In accordance with the multicollinearity problem during computation caused by the beacon nodes used for location estimation which are close to be in the same line or same plane, two solutions are proposed in this dissertation:the geometric analytical localization algorithm based on positioning units and the localization algorithm based on the multivariate analysis method. The geometric analytical localization algorithm based on positioning units analyzes the topology quality of positioning units used to estimate location and provides quantitative criteria based on that; the localization algorithm based on the multivariate analysis method uses the multivariate analysis method to filter and integrate the beacon nodes coordinate matrixes during the process of location estimation. Both methods can avoid low estimation accuracy and instability caused by multicollinearity.3. In accordance with the low coverage problem under certain circumstances, this dissertation has eliminated the multicollinearity affection by using the dimension reductive multivariate analysis method and using a feasible weighted least-square method to solve the heteroscedastic problem caused by error accumulation in practical condition based on the multivariate analysis method. The dimension reductive multivariate analysis method used can not only eliminate the multicollinearity affection, but also reduce noise at the same time. The algorithm is more suitable for the practical deployment situation because the residual is adopted as the weight of the weighted least squares during actual computation, which provides higher localization accuracy.4. In accordance with the problem that the traditional trilateral or multilateral estimation localization method is highly dependent on the proportion of beacon nodes and the measurement accuracy, an algorithm based on kernel sparse preserve projection (KSPP) is proposed in this dissertation. The kernel function is used to measure the similarity between nodes, and the location of the unknown nodes will be commonly decided by all the nodes within communication radius through selection of sparse preservation self-adaptation and maintaining of the topological structure between adjacent nodes. Therefore, the algorithm can effectively solve the nonlinear problem during ranging, and it is less effected by the measuring error and beacon nodes quantity.