NLOS Error Recognition And Suppression Algorithm In Indoor Environment

In recent years,location-based services have been developed in many fields,such as military,agriculture,and industry,due to the development of wireless communication technology and the rise of the concept of the Internet of Things.As intelligent terminals and devices become more popular,improving the accuracy of wireless positioning systems has become a research hotspot.Wireless positioning systems are divided into outdoor and indoor positioning systems based on different scenarios.The technology for outdoor positioning has become mature.However,in indoor scenarios,the complex environment with various obstacles and multipath effects makes the indoor positioning effect less ideal in the Non-Line-of-Sight scene.Therefore,current research focuses on reducing the interference of Non-Line-of-Sight error in the complex indoor environment.Indoor positioning algorithms in the Non-Line-of-Sight environment are studied to meet the requirements of indoor high-precision positioning in this paper.The main research contents and results are as follows:(1)To address the issue of Non-Line-of-Sight error in indoor positioning,an improved suppression algorithm based on the traditional quadratic programming algorithm is proposed in this paper.Specifically,a quadratic programming optimization method based on the geometric constraint is presented,which eliminates the need for prior information of the measurement.By utilizing ranging information,the proposed objective function processes the data and obtains accurate coordinates of the measured nodes and a detailed complexity analysis of the proposed algorithm is carried out.The experimental results demonstrate that the proposed algorithm outperforms the other similar methods in the poor indoor propagation environment.(2)To address the problem of existing both Line-of-Sight and Non-Line-of-Sight data in the process of signal propagation in the indoor environment,a combination of Convolutional Neural Network and Long Short-Term Memory Network is adopted to identify and classify the Line-of-Sight and Non-Line-of-Sight data.Additionally,an improved residual weighting algorithm is proposed,and the classified data is used for location estimation through the proposed residual weighting algorithm.The experimental results show that the localization effect using the classified data is better than the direct localization effect using the residual weighting algorithm,and the computational cost is lower.(3)In order to address the problem of large data errors caused by obstacle interference and multipath effects in the process of signal propagation in the indoor environment,an improved localization algorithm based on positive semidefinite matrix is proposed,which further optimizes the objective function based on the original positive semidefinite matrix.This method directly utilizes Time of Arrival based ranging data for localization without prior location information.Experimental results demonstrate that the proposed localization algorithm has a better positioning effect and higher positioning accuracy than the currently used positive semidefinite method in the same noise environment.Furthermore,the Ultra-Wideband system is used to test the algorithm’s performance in the actual environment,and the experimental results demonstrate that the algorithm performs well in such environments.