Research On Indoor Cooperative Localization Technology Under Waveform Division Multiple Access Ultra Wide Band System

With the development of modern industrial technologies such as 5G communication technology,Internet of Things,and the Internet,indoor high-precision positioning technology has attracted widespread attention from industry and academia.In the indoor environment where the satellite signal of the global positioning system is not covered,the ultra-wideband signal has been widely used due to its advantages of high time resolution and low complexity.Waveform Division Multiple Access(WDMA)based on orthogonal pulse waveform has the advantages of high transmission rate and anti-interference ability in multi-user situations.Therefore,the WDMA-UWB system is an indoor positioning system with practical value combining the advantages of both.Firstly,orthogonal ultra-wideband waveforms are generated based on the design of WDMA-UWB system,and the experimental simulation shows that the requirements of signal used in WDMA-UWB system can be satisfied.Based on the WDMA-UWB system,a cooperative positioning network model based on Bayesian theory is established.Aiming at the problem of ranging error caused by non-line-of-sight propagation,the error mitigation method based on convolutional neural network is studied.The simulation proves that it can effectively reduce the influence of ranging error on positioning accuracy in the positioning system.Secondly,aiming at the problem of nonlinearity and edge posterior probability of measurement state in static Bayesian cooperative positioning network,a belief propagation algorithm based on improved weight Sigma point filter is proposed.The simulation verifies effectiveness of the static node location algorithm.Finally,aiming at the problem that the measurement state of the sensor and the motion model of the node in the dynamic Bayesian positioning network have large linearization error and the calculation of the state matrix is complex,a joint algorithm based on improved weight Sigma point and incremental smoothing is proposed and verified by simulation.The algorithm effectively reduces the root mean square error of the dynamic node position estimation compared with the traditional algorithm.