High Precision Indoor Visible Light Localization Algorithm

White LED is considered as a new generation of green lighting technology due to its advantages of high luminous efficiency,long service life and fast response.Visible light communication based on LED technology has the advantages of low installation cost,no electromagnetic interference,high security and privacy.It quickly becomes a research hotspot in the field of wireless communication.Besides,indoor visible light positioning based on LED has been widely studied.In this paper,the fundamental principles of indoor visible light location algorithms are analyzed,including received signal strength(RSS),time of arrival(TOA),time of arrival(TDOA)and angle of arrival(AOA).In view of the shortcomings of the existing methods,the corresponding improvement methods and measures are proposed from the following aspects.The positioning accuracy of indoor positioning algorithm is effectively improved.Aimed at the problem of low accuracy of traditional positioning algorithm,the RSS and TDOA localization methods based on kalman filter are proposed.In this method,the position estimation obtained by traditional positioning algorithm is taken as the initial value,and the location result is estimated twice by kalman filter,so the positioning accuracy can be effectively improved.In order to solve the problems of poor stability and low positioning accuracy of RSS algorithm,a hybrid positioning method based on RSS/AOA and an improved positioning method based on distance iterative weighting are proposed.In this hybrid localization method,distance estimation and angle information are obtained by RSS signal transmission model and location spatial geometry to construct a hybrid localization objective function.The generalized trust region subproblem method is used to solve the position coordinates of the nodes.Meanwhile,an improved distance iteration weighting method weights the estimated distance according to the characteristics of short visible light transmission distance result in high positioning accuracy.It effectively solves the problem of low positioning accuracy caused by large estimation distance error and improves indoor positioning accuracy.Aimed at the problems of large error and non-convergence of TDOA algorithm,an indoor hybrid positioning method based on TDOA/AOA is proposed.In this method,the distance estimation difference and angle information between the positioning node and the LEDs are used to construct the hybrid positioning objective function,and the coordinates of the positioning node are solved by the second-order cone programming(SOCP)method.The problem that SOCP can't locate the target outside convex hull effectively is solved by introducing penalty term.This method not only improves the positioning performance effectively,but also improves the robustness of the positioning system.Due to the low accuracy of traditional fingerprint localization algorithm affected by environmental factors,a fingerprint localization method based on deep neural network is proposed.This method combines the neural network and fingerprint localization.The RSS and AO A information are obtained according to the RSS signal transmission model and the spatial geometric relationship of localization to constructs fingerprint database.The neural network model is trained by fingerprint database.And then the relationship between the fingerprint data and the position coordinates is obtained to realize localization.This method uses two kinds of signal characteristics as fingerprint data to solve the problem of low accuracy of fingerprint location algorithm.