Optimization Of Visible Light Indoor Positioning Algorithm Based On RSS

With the rapid development of the times,the demand for wireless positioning is increasing.The lack of frequency bands,limited application areas,and expensive equipment are the shortcomings of traditional wireless positioning,which have gradually failed to meet the communication requirements of people.Therefore,in order to solve this problem,many domestic and foreign researchers in the communication field have turned their attention to emerging communication technologies such as optical communication.As the importance of traditional lighting sources in the lighting market is declining,light-emitting diodes(LED,Light-Emitting-Diode)are widely used.At the same time,compared with traditional sodium lamps,incandescent lamps,fluorescent lamps and other light sources,the power consumption of LED is lower.The luminous intensity is not inferior to the traditional lighting source,and the service life is greatly improved,which is also the advantage of LED.Based on the above,a new positioning method combining optical communication technology and LED has been proposed by some scholars,that is,LED visible light communication technology(VLC,Visible-Light-Communication).Therefore,the development of visible light indoor positioning is promoted by VLC.The working principle and system construction of indoor positioning in visible light are explained in detail,and a model of the actual environment is constructed.The main innovations of this article are as follows:(1)Based on the indoor visible light positioning of RSS(Received Signal Strength)fingerprints,the fingerprint database cannot be quickly updated by this method under the gradual change of the environment,so a lot of costs will be required.According to the sparse characteristic of the matrix constructed by the fingerprint database,the method of updating the fingerprint database can be transformed into a method of filling the matrix.Next,the issue of noise and data continuity impact on accuracy needs to be paid attention to.At the same time,local correlation restrictions are introduced into the traditional matrix filling model.Finally,in the offline stage of the fingerprint database positioning algorithm,a matrix filling model with smoothing restrictions will be used in it;In the online matching stage,the fast calculation speed and the small amount of calculation are the advantages of the segmented orthogonal matching pursuit algorithm(St OMP),which is suitable for fingerprint library reconstruction.The impact of choosing different thresholds on the matching results of the St OMP algorithm is needed,so an algorithm that automatically adjusts the threshold in time is necessary.Due to the idea of proportional,integral,and differential methods,we calculate the current matching error of the St OMP algorithm.The next threshold change value is calculated by this error.Finally,the fingerprint database will be matched and reconstructed in the online stage.(2)Under the condition of sudden changes in the environment,the RSS value fluctuation of the target positioning technology based on RSS fingerprints leads to the decrease of positioning accuracy.Therefore,the obtained intensity value is changed into the following two parts: one is the fluctuation component caused by the positioned object,and the other is various interference noises caused by the external environment.The linear transition model of the state Markov chain is used to extract the disturbance component in the RSS value.A dynamic time warping algorithm is used to match the real-time RSS target disturbance component with the fingerprint library to achieve target positioning.Experiments show that it can effectively reduce the impact of environmental noise and improve the robustness of the positioning system while ensuring positioning accuracy.