Research And Verification Of Indoor Visible Light Depth Positioning Network Based On Received Signal Strength And Phase Difference Vector

With the popularization of concepts such as smart home and the Internet of Everything and the realization of related technologies,indoor positioning technology,as a basic service,is increasingly in demand,and traditional indoor positioning technology has problems such as high cost or insufficient accuracy.Indoor VLP technology,as a new type of indoor positioning solution,naturally has the advantages of environmental protection,low power consumption,and low cost,and has become a research hotspot in indoor positioning technology.However,factors such as noise,ambient light interference,and optical signal transmission medium instability in space and time domains affect the positioning accuracy and stability of VLP.This paper fully studies and analyzes the theoretical basis,characteristics and existing technical solutions of indoor visible light positioning.Combining with the channel estimation method based on image noise reduction,a new fingerprint construction method and noise reduction depth positioning network are proposed.The main research content of this paper is as follows:(1)This paper systematically analyzes the characteristics of the indoor VLP model,and comprehensively considers the advantages and disadvantages of different algorithms in terms of positioning accuracy and complexity,and proposes a method based on the RSS vector or arrival phase FDNet positioning method of difference vector.(2)The core idea of the FDNet positioning method is to use the RSS or arrival phase difference information received by four adjacent PDs at the receiving end to form a multi-channel "fingerprint image",and then use the improved noise reduction network to process this fingerprint image.Noise reduction processing to obtain fingerprint information with a relatively high signal-to-noise ratio,and finally use the BP network to realize the mapping of the fingerprint library to complete the positioning.Simulation results show that the average positioning error of FDNet based on RSS vector is 1.93cm,and the positioning error of FDNet based on arrival phase difference vector is 1.43cm.The positioning accuracy and stability are better than traditional optimization algorithms.(3)Based on the positioning method proposed above,the hardware implementation and performance verification of it are carried out by using Field Programmable Gate Array(Field Programmable Gate Array,FPGA).In the process of FPGA hardware implementation,aiming at the deployment and implementation of FDNet on FPGA,a data quantization method with low precision loss is adopted to carry out targeted parallel optimization and pipeline design.In addition,this paper also introduces the design scheme of the transmitting and receiving modules of the indoor VLP system in detail.Finally,the verification of the hardware system was carried out,and the average positioning error of the test results of the hardware system was 3.47cm.The results show that the indoor VLP system realized by the FDNet method proposed in this paper has high positioning accuracy and system stability.