Target localization based on weighted total least squares in underwater acoustic networks

Most existed target localization methods in underwater sensor networks assume that the locations of sensor nodes are accurate as references to infer target location. However, in practice especially in underwater acoustic networks, the self-localization errors are inevitable. In this paper, considering the presence of both self-localization errors and measurements noises, a target localization model using time difference of arrival is constructed and then the localization problem is formulated. In order to solve the problem, we propose a TDOA target localization algorithm based on weighted total least squares in an iterative fashion. The proposed model takes into account the prior information of measurement noise variance and self-localization error variance which improve the performance. Simulations are conducted to validate the effectiveness of the proposed method. The results show that the proposed WTLS-based target localization method has superior performance compared with the existed method.