Research On Indoor Positioning Algorithm Based On Visible Light

With the continuous development of wireless communication technology and mobile terminal equipment,people's demand for Location Based Services(LBS)is increasingly strong.So it is particularly important to use positioning technology to navigate individuals indoors.The indoor positioning system with visible light communication technology as the core is gradually becoming a research hotspot because of its advantages of simple equipment,low cost and simple operation.This paper first analyzes the indoor visible light positioning channel model,including direct model and noise model.At the same time,several common indoor visible light positioning methods are introduced,and the positioning algorithm based on RSSI is emphasized.In view of the general lack of experimental verification of visible light positioning algorithm research,a new set of indoor visible light positioning system was designed.And the paper focuses on the design ideas of the hardware circuit and the selection of transceiver devices.In addition,based on the constructed experimental platform,the parameters of the Lambertian radiation model were measured.Considering that there is always ambient light interference in the actual positioning scene,and the ambient light will cause the RSSI value to deviate from the actual value.As a result,the calculated distance between the transceiver and the receiver is smaller than the actual distance.Therefore,a method of environmental error offset correction is proposed based on the least square positioning algorithm.At the same time,the positioning performance under different transmission power was explored.Experiments show that the method can achieve accurate positioning based on indoor visible light in a natural environment.The traditional ranging positioning method needs to know the position coordinates of the LED and establish its own propagation model in advance.And the built model is sometimes susceptible to interference such as background light and wall reflection.To further solve this problem,an improved location fingerprint positioning method is proposed.Firstly,in view of the shortcomings of the existing location fingerprint library construction method,a fusion fingerprint reconstruction IALM-OK(Inexact Augmented Lagrangian Multiplier-Ordinary Kriging)algorithm is proposed,which reduces theworkload of fingerprint collection while ensuring reconstruction accuracy;In the online stage,in order to reduce the traversal time of matching and improve the efficiency of matching algorithm,the idea of clustering region is introduced.Aiming at the problems that the initial clustering center of traditional K-means algorithm is randomly selected and the K value of the cluster number needs to be set in advance,an improved K center clustering algorithm is proposed.At the same time,WKNN(Weighted K-Nearest Neighborhood)algorithm is used to achieve positioning,which reduces the time consumption of positioning and improves the accuracy of positioning.