The positioning accuracy of the TDOA location system is mainly influenced by the time difference measurement and the sensor geometry distribution. For the real-time positioning, due to the mobility and the complexity caused by the environmental goal control, the reasonable, efficient and accurate determination of the measured time delay difference is a plagued issue for researchers. To solve this problem, this paper proposes a method to estimate the error signal delay discrimination based on characteristic parameters. The Newton iterative search algorithm is improved to search for the initial positioning and the end condition. Firstly, the signal characteristic value information is used to establish credibility, then the initial position of the source is found through three sensors of high reliability, its results are taken as the initial values of Newton iteration to avoid the complicated calculation of the traditional method. Finally, with the reliability of each sensor as the weight of the Newton iteration termination condition, the end condition is more reasonable, more close to reality. The experiment shows that this algorithm has high precision and strong robustness.

gross time delay error  /  Newton iterative  /  sensor's credible degree  /  positioning accuracy