Research On Localization Algorithms Under NLOS Wireless Propagation Environment

In the past two years,with the rapid development of mobile internet and mobile applications,location-based services(LBS)become more and more popular.As a result,a higher accuracy in a more complex and harsh wireless environment is requried in wireless location under various network standards.And the non-line of sight(NLOS)issue becomes a key problem to be solved.The NLOS error happens when there are blocks between transmitters and receivers,which make radio signal propagation delay excessivly by reflection,refraction and some other reasons.In these cases,the conventional LOS location algorithm will fail to work.That's because their models do not take NLOS into consideration,which results in the solution to the equtions is far away from ture position.In this thesis,the radio propagation models on NLOS / LOS environments were studied.Firstly,we introduced several statistical models and probability distribution of NLOS error,and analyzed the impact of NLOS in the conventional localization algorithms.By comparing some differences of geometric parameters in LOS and NLOS environments,the traditional weighted least square NLOS suppression algorithms were extended to the iterative ones.We discussed the performance of the algorithm via different equations and residual parametric by both theory and simulations.The results show that the IWLS algorithm can effectively improve positioning accuracy rather than conventional algorithms under the same conditions.In the end,the thesis introduced the range shrink algorithm,and analyzed the limitations of the existing algorithms.We propose some new parameters to indicate the degree of LOS reconstruction(the area of feasible region,e.g.).Simulations show this kind of algorithms can work effectively under the DDM-NLOS environment.