Robust Time-Based Localization In NLOS Environments

With the increase of people's requirements for the quality of localization services,satellite localization systems have been unable to meet the needs of user groups for a large number of specific locations.Wireless network localization technology can not only achieve target localization requirements for specific scenarios,but also solve the problem that satellite systems cannot perform accurate indoor localization.The time information-based localization method has the characteristics of fast speed and high accuracy,and has been widely used in engineering practice.However,in practical environments,it is difficult to maintain strict clock synchronization between nodes in a wireless network.At the same time,when the signal propagates through obstacles such as walls,refraction and diffraction occur,which is called non-line-of-sight(NLOS)propagation of the signal.The NLOS error caused by this propagation method will seriously affect the target localization accuracy.So how to overcome the impact of these negative factors on the accuracy of target localization is the main research content of this paper.The main contribution of this paper can be summarized as follows:1.Analysis of the existing robust least squares(RLS)methods for target localization under NLOS environments reveal that there are two shortcomings that seriously affect the accuracy of target localization.One is that the upper bounds of NLOS errors are too large in the derivation of formula model,which leads to too large regularization term in the general least square(LS)problem,and thus leads to the decrease of solution accuracy.The other is the application of triangle inequality in the method,which cannot guarantee the accuracy of the problem.To overcome the shortcomings of existing RLS methods,this paper proposes two new RLS methods based on TDOA.First,this paper reduces the upper bounds of the NLOS errors to a reasonable range by jointly estimating the NLOS error and the target source position in the reference path.Secondly,in order to avoid the use of triangular inequality,two methods of target localization solution are derived by introducing the "balance parameter" and the identity transformation of the measurement model.Both methods utilize S-lemma to equivalently transform the robust maximized part of the objective function of the original RLS problem.2.Clock asynchronous target localization in a NLOS environment is a new research topic.Changing the signal sending and receiving method in the asynchronous network can save energy consumption for the target source and achieve the target self-localization function,but it will lead to the introduction of some parameters that affect the problem solving in the TOA(Time-of-Arrival,TOA)measurement model.Selecting the reference anchor node and converting the TOA measurement model to the TDOA measurement model can eliminate these parameters,and thus the RLS problem of jointly estimating the target source position and the reference path NLOS error can be constructed.The RLS problem is converted into an SDP problem and solved.The obtained result of target location has high accuracy and good robustness to the NLOS error in non-reference path measurements.