Research On Key Technologies Of Imperceptible Visible Light Communications

As a new wireless transmission technology,Visible Light Communications(VLCs)technology has attracted much attention and developed rapidly due to its wide spectrum resources,electromagnetic interference-free,safety,combining lighting and communication.VLC takes advantage of visible light with the wavelength of 380 nm to 760 nm for communication,which may produce glare and flicker.These uncomfortable feelings result in VLC cannot be applied in the illumination restricted scenario(such as dim bedroom,cinema).Therefore,VLC is not an "Always-on" communication technology compared to RF-based communication technologies such as Wi Fi.The human eye discomfort caused by communication light is one of the obstacles to the development of VLC.In this thesis,the concept of imperceptible VLC(iVLC)is proposed,key technologies of iVLC are researched,and the "Always-on" VLC system is realized.In general lighting communication scenarios,downlink VLC can achieve no flicker effect on human eyes by combining lighting and communication technologies.However,for illumination restricted scenarios,VLC communication can cause glaring problems.Therefore,the current situation is described as colloquial,namely: the light goes out,communications of VLC are cutoff.Uplink of VLC can also cause glaring problems.Currently,the common solution is to use RF and infrared as uplink communication links.In 2016,a research team at Dartmouth College in the United States proved that VLC can be achieved in seemingly dark environments,which they called "Dark VLC".Inspired by their work,this work explores the theory and technology to solve the "glaring" problem of VLC,aiming at the realization of iVLC.Unlike the flickering sensation,the glaring sensation results from the difference between the brightness of the luminous body and the brightness of the environment being higher than the mental threshold that the human eye can perceive.In order to satisfy the communication requirement that human eyes cannot perceive the light in visible band,this work introduces the theory of psychophysics and proposes the just imperceptible difference(JID)which can measure the imperceptibility of human eyes.Thus,the measurement problem of glaring sensation is solved.Based on JID,the upper bound of iVLC average transmitting power which is a necessary condition for iVLC is studied in this work.Aiming at the problem of high difficulty in synchronization and detection caused by low average communication power of iVLC system,the synchronization and detection technologies which are suitable for iVLC system are studied in this work.To further enhance the synchronization and detection capability of the iVLC system and enable the iVLC system to have channel equalization capability,deep neural network(DNN)is introduced in this work and a multi-quadric kernel and DNN based hard-max pulse position classifier(MQK-DNN-HPPC)is proposed.The main research contents and innovation work of this thesis are as follows.(1)In view of the glaring problem of VLC system,the psychophysical theories are introduced into iVLC system design in this work,so as to realize VLC that cannot be perceived by human eyes,and make VLC have the characteristics of "Always on".Based on the theory of psychophysics,JID which can measure the imperceptibility of human eyes is proposed.JID is the maximum stimulus intensity difference between two stimuli of the same kind that people cannot distinguish.In this work,the two-dimension channel characteristics of iVLC are characterized by communication channel and perception channel.Based on JID,the upper bound of iVLC system average transmitting power is derived in this work,which is limited by the imperceptibility requirement of human eye.In an environment with arbitrary illumination,VLC can realize imperceptibility of human eyes during communication when meeting this requirement.Based on the communication channel model of iVLC system,the lower bound of the average transmitting power of iVLC system limited by the communication reliability is derived.Pulse position modulation(PPM)is employed to reduce the average power of the transmitted signal while preserving the detectable pulse power for the receiver,so as to ensure that a feasible region exists between the upper and lower bounds of the average transmitting power of iVLC system.Using the traditional VLC receiver noise parameters,the theoretical feasibility of implementing iVLC is verified by simulation,and the key factors affecting the performance of iVLC system are studied,which provides guidance for the design of iVLC system and its components.(2)In illumination restricted scenarios,iVLC needs to keep its average transmitting power at a low level,which makes synchronization and detection more difficult.In this work,frame synchronization,symbol synchronization,and slot synchronization methods are designed for the PPM iVLC system.It includes the design of frame synchronization sequence for data frames with different PPM order by using Barker code characterized by strong autocorrelation.Taking advantage of the characters of PPM symbol and time slot,the symbol synchronization method and time slot synchronization method are designed respectively.Simulation results show that the proposed synchronization method can enhance the robustness of iVLC system against timing offset.Aiming at the difficulty of detection of iVLC low average power signal,the detection capability of three detection methods that can be used in iVLC system are simulated and analyzed in this work.(3)In order to further improve the synchronization and detection capability of iVLC system and enable iVLC to have channel equalization capability,DNN was introduced in this work and MQK-DNN-HPPC which is suitable for PPM modulated iVLC system was proposed.MQK-DNN-HPPC is a joint synchronization,detection and equalization method of machine learning which is constructed into an overlapping multi-layer structure.This overlapping multi-layer structure can increase the utilization rate of training data and speed up training speed.In order to highlight the characteristics of training data and improve the performance of neural network,MQK layer is introduced in MQK-DNN-HPPC.Simulation results show that MQK-DNN-HPPC can improve the communication reliability,synchronization performance,communication rate and feasible range of transmitting power of iVLC system.In this thesis,a new theory is proposed to solve the VLC blinding problem,and the key factors to realize iVLC(such as average transmission power,detection method,signal duty ratio,ambient light intensity,etc.)are studied.Suitable synchronization,detection and channel equalization methods are designed for iVLC.The research results of this work can provide guidance for iVLC system design and optoelectronic device design.