Study On Indoor Localization Algorithm Based On The PSS Of Long Term Evolution(LTE)

As the most mature positioning system,Global Positioning Satellite Positioning System(GNSS)is mainly used in relatively open environments.In the indoor environment,the walls and obstacles are relatively dense,and the signal transmission is severely blocked,causing the positioning system to “lost the stars”.Unable to complete targeting.Currently,the commercial LTE(Long Term Evolution)mobile communication system provides signals such as PRS,PSS,and CRS,which can be used for positioning.However,the propagation of LTE signals in indoor environments is highly susceptible to NLOS(None-Line-of-Sight)and multipath fading,resulting in extremely large time of arrival(TOA)estimation errors,resulting in reduced indoor positioning accuracy,and Operators usually do not send PRS signals.Therefore,in view of the above problems,this paper develops an accurate TOA estimation using the PSS primary synchronization sequence to provide an indoor positioning scheme based on the LTE mobile communication system.The scheme is divided into three parts.Firstly,the channel impulse response and multipath delay of multipath channel are estimated by compressed sensing.The compressed sensing channel estimation algorithm based on LTE primary synchronization sequence is designed and implemented.The estimated multipath channel delay and impulse response are obtained.Accurate TOA value.Simulation experiments verify the effectiveness of the algorithm.Secondly,in order to improve the estimation accuracy of TOA in NLOS scene,and using integrated learning to have the ability to speed up training time,a LOS/NLOS discriminator based on AdaBoost(Adaptive Boosting)algorithm is proposed to effectively improve the non-line-of-sight discrimination efficiency.Numerical simulations show that compared with the traditional non-line-of-sight discrimination algorithm,the computational complexity of the proposed algorithm is greatly reduced under the condition of obtaining approximate classification accuracy.Finally,the paper refers to the classification idea of linear regression,uses the one-dimensional convolutional neural network to train the training data with error labels,establishes the error skewness classifier,and proposes the LTE-TOA estimation algorithm based on neural network.Simulation results show that this method can reduce the TOA estimation error and improve the indoor positioning accuracy compared with traditional algorithm.