Research On UWB Based Indoor Non-line-of-sight 3D Positioning Technologies

In recent years,location-based services developed by wireless sensor networks have attracted extensive attention,and are widely used to find and guide emergency personnel,indoor location games,virtual reality and other fields,in which location accuracy is the key problem.Most of the location-based services are based on the global positioning system to obtain accurate location information.However,due to various obstacles and the attenuation of GPS signal through the wall,it can not provide accurate location services in the indoor environment.Although a lot of work has been done to overcome this problem,and many technologies have been used to obtain accurate location in indoor infrastructure,indoor location is still a challenge area to be explored.Under this background,this paper studies the indoor non-line-of-sight(NLOS) 3D positioning technology based on ultra wideband(UWB).Firstly,this paper summarizes the research background,significance and current situation of indoor positioning technology,discusses and analyzes the advantages and disadvantages of common indoor positioning technology,and focuses on the basic theory and technical characteristics of UWB indoor positioning technology.Theoretical analysis shows that UWB signal has the characteristics of strong penetration,strong anti-interference and strong multi-path resolution,which is unique in indoor positioning system Advantage.Then the location estimation algorithm based on UWB is studied and analyzed,and the location algorithm based on TOA and the location algorithm based on TDOA are tested.The test results show that the three-dimensional location accuracy of these two algorithms fluctuates greatly in NLOS environment,and the location accuracy is low,which fluctuates up and down at 23cm and 47cm,20cm and 45cm respectively,which is not suitable for the three-dimensional location in complex environment.Then,this paper proposes a location algorithm based on the generative adversarial networks(GAN).Based on the theory of generative adversary network,this algorithm improves the identification network and generative network,optimizes the loss function of generative network,and finally tests the algorithm.The experimental results show that the algorithm is based on GAN in NLOS environment.The three-dimensional positioning error of the positioning algorithm against the network is about 20cm,and the positioning accuracy fluctuates little and the stability is high.At the same time,three positioning base stations are tested.The results show that for three positioning base stations,the algorithm still has good positioning accuracy and stability in complex environment,and the positioning error is about 25cm.Finally,a set of UWB indoor positioning system is built,the hardware modules of base station and label are introduced,and the upper computer software of the system is designed based on Java Script and node.js technology,which provides a visual interface and is easy to operate and use.