| In recent years,location based services(LBS)have been rapidly popularized in navigation,monitoring,message push,shopping guidance and other applications,producing a huge demand for accurate indoor positioning technology.At present,the mainstream positioning method is satellite positioning.But indoors,satellite signals are blocked by buildings.Indoor location technology based on magnetic field information has gradually become a research hotspot because of its advantages such as no additional location facilities,high positioning accuracy and relative stability.As the magnetic field values in a single position are very limited,in order to improve the precision of the magnetic field,the continuous magnetic sequence collected in the walking process should be used as the characteristic value of the magnetic field to improve the recognition degree.This kind of indoor location method based on magnetic field sequence matching has attracted more and more attention.In this paper,the performance evaluation framework of an indoor positioning system is designed,and typical three types of typical scenes with different magnetic properties are selected,and the performance of the existing algorithms in different scenarios is simulated and verified by experiments.Our analysis obtained the following three conclusions.First,through the analysis of the complexity of the matching process,it can be seen that the positioning accuracy has a great dependence on the step of the sliding window search,and it decreases rapidly with the increase of the step.Second,the positioning precision can be improved by increasing the matching length within a certain range.However,because of the existence of random error,the positioning accuracy can not increase indefinitely with the increase of the matching length,and the matching length of the magnetic field in the same position increases,and the location result is reduced.Third,magnetic field feature selection also has great influence on positioning accuracy.The application of feature dimension expansion in DTW matching can effectively improve the positioning accuracy.Based on the systematic analysis of the existing methods,the paper further improves the shortcomings of the existing algorithms.The following three aspects are optimized.At the same time,while taking into account the complexity of operation and positioning accuracy,it is proposed to search for the starting point based on the alternative sequence,which has higher search efficiency,and has higher performance in a relatively large range of indoor magnetic fields.Secondly,introducing adaptive matching length can improve the location of magnetic field matching and reduce the computational complexity of matching.In the end,the adjacent position information is introduced to match the position of the matching trajectory,and the hidden Markov model is applied to eliminate the large scale error of the matching process,and the limitations of the use are analyzed.Simulation and experimental results show that the new improved method improves the computational complexity and positioning accuracy in a magnetic field stable scenario. |
