Network Node Substep Localization Algorithm Research

Localization of nodes is a key technology for application of wireless sensor network and has been studied extensively in recent years. However, there are still many obvious deficiencies and shortcomings in this localization technology in wireless sensor network. First, there's always a lack of node localization structure with high positioning accuracy and low computational burden in the literature since the existing methods can't weight the objective function, or require the N-dimension search, or can't utilize all of range measurements. Second, the position error of anchor nodes is not considered in wireless sensor network. Third, the localization algorithm in wireless sensor network is lack of efficient learning localization algorithm.The objective of this thesis is to solve the problems of the existing sensor location methods which have the low positioning accuracy and high computational burden, can't consider the position error of anchor nodes, and the performance analysis is only suitable for special situations. The novel node localization structure with high positioning accuracy and low computational burden is proposed in this thesis first. Then several node location methods based on the proposed localization structure are derived in this thesis for both LOS and NLOS environments.The primary contributions of this thesis are summarized as follows:(1) This thesis proposes a novel node localization structure with high positioning accuracy and low computational burden. The proposed structure solve the problems of the existing node localization structure which can't weight the objective function, or require the N-dimension search, or can't utilize all of range measurements; this study provides a foundation for further research on node localization methods.(2) TOA/AOA based WSN localization algorithm in LOS environment and TDOA/AOA based WSN localization algorithm in LOS environment by the proposed node localization structure are proposed. Since the proposed method with high positioning accuracy can obtain the localization of nodes cost function by the useful geometric characteristics, covariance matrix of range measurements and the position error of anchor nodes, and the closed-from solution for the localization of nodes cost function is available by the least square method.(3) TOA/AOA based WSN localization algorithm in NLOS environment and TDOA/AOA based WSN localization algorithm in NLOS environment by the proposed node localization structure are proposed. Since the proposed method with high positioning accuracy use robust cost function that is not sensitive to the gross error instead of mean square cost function, and the robust cost function can restrain the gross error by weighted and the closed-from solution for the robust cost function is available by the least square method.(4) This thesis proposes an improved LS-SVM based location algorithm to solve mobile location problem in a NLOS environment. The proposed method can obtained the rough position estimation of blind nodes by LS-SVM location. Moreover, steepest descent method is used in the proposed method to iterative search the optimal position estimation of blind nodes. The proposed method can improve the positioning accuracy by using all the range measurements among the nodes.