Research On Indoor Location Algorithm Based On WiFi And Bluetooth Fusion

With the rapid development and popularization of Location Based Services(Location Based Services,LBS),accurate and convenient indoor positioning services have attracted more and more public attention.In indoor scene,WiFi location algorithm based on location fingerprint is widely recognized as an indoor location method.How to reduce the interference caused by multi-path and attenuation caused by wireless signal propagation in location fingerprint algorithm,how to solve the problem that Received Signal Strength Indicator(Received Signal Strength Indicator,RSSI)signal fluctuates greatly,the maintenance cost of fingerprint database is high in the later stage,and the location of users in indoor environment can be tracked accurately has become an urgent problem to be solved by scholars nowadays.To this end,this paper mainly does the following work:1.Because of the diversity of indoor positioning methods and algorithms,the previous indoor positioning methods were studied.This paper analyses the differences between the computational complexity and the positioning accuracy of several positioning methods,and chooses suitable calculation methods and physical measurement methods according to the application requirements of this paper.The superiority of the deterministic algorithm is verified by actual positioning experiments and MATLAB simulation.The experiment shows that the deterministic algorithm has certain advantages in positioning time,and the positioning accuracy of K-nearest Neighbor(K-nearest Neighbor,KNN)algorithm is slightly better than that of bayes algorithm.proving the reliability of deterministic algorithm in indoor positioning.2.Aiming at the interference of multi-path and path attenuation caused by signal propagation,this paper carries out the acquisition experiment of WiFi signals in indoor environment,focusing on various factors that may reduce the accuracy of positioning.The experiment shows that there is a problem of terminal heterogeneity in the acquisition stage.The RSSI values collected are filtered and processed,and the filtering results are obtained by MATLAB simulation.he experiment shows that kalman filter is better than mean filter and Gaussian filter in system stability,which confirms the reliability of Kalman filter for RSSI value preprocessing Because the signal propagates to a certain point and fluctuates in a certain range during the construction stage of fingerprint database,logarithmic attenuation model is used to optimize the parameters of actual data.3.A region weighting algorithm based on the fusion of WiFi and Bluetooth is proposed,which can effectively reduce the influence of multipath interference in signal propagation.In the off-line analysis stage,we use the multi-lateral measurement of WiFi and Bluetooth signal strength to get the appropriate location area,and use the area to set up confidence,then get the fusion position information of WiFi and Bluetooth by fusion weighting,which can effectively fuse the location of WiFi and Bluetooth.Because there are errors between the fusion location results and the pre-processed position coordinates,and the errors are regionalized,the moving least squares interpolation(Moving Least Square Interpolation,MLSI)method is used to fit the fusion errors and obtain the fitting function of the errors.Compared with the global unbiased estimation of fusion error,it is more suitable for indoor fusion error model.Compared with cubic spline function,Gauss weighting function has more shape parameters and is more flexible.By adjusting shape parameters,the data can be updated periodically.The reliability and adaptability of Gauss weighting function are verified by MATLAB simulation experiments.Aiming at the problem of fingerprint database maintenance in the later stage,the reference nodes of error fitting are updated regularly by using crowdsourcing nodes with better customer feedback,which greatly reduces the cost of survey and maintenance in the later stage.Experiments show that the algorithm has a great improvement in positioning accuracy and stability.