Research on TDOA Localization under Unknown Signal Propagation Speed
Localization for the time-difference-of-arrival(TDOA)measurement under unknown signal propagation speed is studied,and two localization methods based on joint signal propagation speed and target location estimation are proposed. The first method is the two-step weighted least squares method. In this method,an initial weighted least squares estimation of unknown parameters is obtained with-out considering the relation between parameters. To improve the estimation performance in the first step,the second step is considered to transform the relationship between the parameters in the first step into a standard generalized trust-region subproblem,thus higher preci-sion estimation performance is obtained. The second method is the positive semidefinite relaxation method,which constructs the nonlin-ear non-convex weighted least squares problem, and then a semidefinite programming optimization problem is formulated through utilizing semidefinite relaxation. The result of simulation shows that the two proposed methods are very close to the Cramer-Rao lower bound under Gaussian noise when the noise is not too loud.
localizationSemidefinite relaxationtime difference of arrivalMaximum likehood estimationtwo-step weighted least squares method