Research On Parameters Estimation Methods For Mobile Terminal Localization

With the evolution of communication technology,Mobile Localization Technology draws more and more attention.However,the location accuracy has become the main challenge that limit the further development of location-based services. Due to the complexity of the wireless channels,noise,multiple-access interference, multipath interference and non line-of-sight would degrade the estimation performance. How to improve the location accuracy in the complex wireless environment has become the critical issue of wireless localization.Based on location-related information such as time delay,angle,attenuation and other parameters,wireless localization mainly be carried out to locate mobile terminal. In the view of the drawback of single parameter estimation,this thesis presents the further study on joint parameters estimation method(TDOA/AOA) in the multipath and noise environment.This paper firstly introduces the classic ESPRIT TDOA/AOA joint parameters estimation method,and then makes analysis in its capability with context of the same delay.However,this algorithm does not fully utilize the information provided by channel response matrix,so the paper puts forward the joint estimation method based on the eigenvalues and eigenvectors.This algorithm estimates the delay and angle for only one eigenvalues decomposition.Thus it takes full advantage of the matrix information and makes the accuracy better.Aiming at the gaussian noise in wireless channel and according to the insensitivity of high-order cumulant to gaussian noise,this paper proposes multipath joint parameters estimation method based on high-order cumulant.It first estimates the delay and attenuation,furthermore,revises the attenuation while estimating the angle.Therefore the method can improve the accuracy obviously.Finally,combining high-order cumulant with the ESPRIT method,the paper brings up separation algorithm.Adopting the high-order cumulant in the estimation of delay and angle,it estimates the angle with ESPRIT method and verifies its effectiveness by the simulation.