Multipath Estimation In The Framework Of Kalman Filter

With the development and application of satellite navigation system,the accuracy of positioning is required higher and higher.Multipath error is one of the main error sources affecting the high precision positioning.It is of great significance to study the multipath error suppression method to improve the positioning accuracy of navigation system.In recent years,the multipath error suppression method based on data processing has attracted much attention due to its advantages of low cost and easy development.The core of this method is multipath parameter estimation.The parameter estimation algorithm of Kalman filter framework is often used for multipath parameter estimation due to its advantages of easy implementation and good estimation performance under Gaussian noise.However,the estimation accuracy of this kind of multipath estimation algorithm will decrease significantly in the non-Gaussian noise environment.In view of this,this paper studies the multi-path estimation algorithm under the framework of Kalman filter under nonGaussian noise,and mainly completes the following work:1.A Generalized Maximum Correntropy Unscented Kalman Filter(GMCUKF)multipath estimation algorithm is proposed to solve the problem that the traditional Minimum Mean Square Error(MMSE)based multipath estimation algorithm is not suitable for the non-Gaussian noise environment.In this algorithm,generalized correntropy is used as the criterion to measure the similarity between the estimated value and the real value,and the optimal estimate is obtained by means of fixed-point iteration.Compared with the Unscented Kalman Filter(UKF)multipath estimation algorithm and the Iterated Unscented Kalman Filter(IUKF)multipath estimation algorithm,the results show that the GMCUKF-based multipath estimation algorithm has higher estimation accuracy in the non-Gaussian noise environment.2.For GMCUKF multipath estimation algorithm,if there is a large observation error in the operation process,the error inversion of diagonal matrix will not be carried out normally.In addition,the algorithm uses fixed noise covariance estimation recursively,and it is difficult to obtain better multipath estimation accuracy.An Adaptive Iterative Correntropy Unscented Kalman Filter(AICUKF)multipath estimation algorithm is proposed.This algorithm defines a cost function based on the correntropy criterion,and derives two scalar weights for adjusting gain,which can effectively avoid the occurrence of GMCUKF numerical problem.On this basis,an adaptive algorithm based on the idea of covariance matching is introduced to realize the adaptive updating of noise covariance matrix,which can ensure that the noise covariance is closer to the real value and further improve the performance of multipath estimation.The comparison results with the UKF multipath estimation algorithm and the Adaptive Unscented Kalman Filter(AUKF)multipath estimation algorithm show that the AICUKF-based multipath estimation algorithm can achieve higher multipath estimation accuracy in the non-Gaussian noise environment.In this paper,based on Kalman filter framework,the entropy criterion is used to improve the UKF algorithm,which solves the problem of the performance degradation of UKF multipath estimation algorithm in non-Gaussian noise environment,and further improves the accuracy of multipath estimation in nonGaussian environment.