Key Technology Research On Signal Design For Massive MIMO Visbile Light Communication Systems

Combining the lighting and communication,visible Light Communication(VLC)gains becomes attractive to many researchers,with the characteristics of the green,safa,rich spectrum.Moreover,in typical illumination scenarios,multi-lamp scene provides a natural MIMO structure.Currently,due to the shortage of radio frequency communication spectrum and the high requirement of users to the communication rate,high-speed transmission and industrial applications become two major research hot spots for VLC.Then,massive MIMO-VLC comes into being,which has three major characteristics: lighting,communications and industrialization.The “industrialization” means that the system should be multi-array,micro-spacing and the transceiver needs to have a good mobility.The combination of the “lighting” and “communication” causes great challenges to the theoretical and technical research.Thus,The main research topic of this paper is the key technology study on the signal design for indoor massive MIMO-VLC systems,under the premise of taking into account the communication measure and lighting requirements.Then,this paper summarizes our subject as three key issues: the theory limitation to system achievable rate,the achievable rate with availability-reliability tradeoff,and the energy-efficient constellation design.The three key issuses is a progressive layers of the process.The main contents of this paper are as follows:In the first part: Theory limitation to system achievable rate.As for indoor massive MIMO-VLC systems,we derive the point-to-point and ergodic channel capacity.First,based on the unique geometry of indoor VLC channel,a general approximation lower limit is proposed for the difference entropy of the received signal,which has a simple relation to the geometric parameters.Based on this,the lower upper limit of point-to-point capacity of VLC channel is deduced.Then,according to the daily people's indoor activities,we model two typical two-dimensional distributions.Based on the point-to-point capacity,we derive the lower bound of the VLC channel traversal capacity.From the simulation results,the capacity bounds have good tightness,and are effective to reveal the relation between the capacity with inter-LED interference(ICI),channel gain and so on.Note that the capacity research in the second chapter is the basis of subsequent chapters.In the second part: Achievable rate with availability-reliability tradeoff.Taking the idea of using channel characteristics to reduce the ICI,to study the diversity-multilexing tradeoff(DMT)and to propose a channel-adaptive spatial modulation with balancing the effectiveness and reliability balance.1.Effectiveness-reliability tradeoff for massive MIMO-VLC systems with illumination constraints.Based on the time-invariant and geometric characteristics of channel,the relationship between validity and reliability is studied deeply,and the optical effectiveness-reliability tradeoff(OERT)is established.Our main work is as follows: Firstly,we propose the new definitions to measure the effectiveness and reliability gain.Moreover,combined with optical power and LED nonlinear characteristics,we introduce the power gain and the spatial gain.Secondly,combining the communication performance and lighting requirement,we establish the optimization problem for OERT.Thirdly,we propose a fast search algorithm to reduce the computational complexity,for the solution to the optimal OERT scheme(OERTS).Finally,the performance of DMT is analyzed from different aspects based on three performance curves.From the simulation results,the OERT not only achieves the tradeoff between communication effectiveness and reliability,but also reflects the influence of LED nonlinear constraints on optical power utilization.Moreover,the optimal OERTS also shows good adaptability to the channel and has direct guiding significance for VLC application promotion.2.Study on the channel adaptive spatial modulation.Based on the DMT study,aiming at the MIMO-VLC with high-correlation or even lack-rank channel,a universal criterion is established for balancing BER performance and spectral efficiency.On the basis of this,a channel adaptive spatial modulation(CASM)scheme is proposed based on the geometrical properties of MIMO-VLC channel.From the simulations,CASM can effectively improve the communication performance.Even if we consider the 10% of the spectral efficiency loss by the channel estimation,it can still achieve better error performance.Moreover,the corresponding constellation set is adapted to the channel,improving the receiver mobility of MIMO-VLC system.In the third part: Energy-efficient constellation design.Taking the idea of transforming uncontrollable ICI into controllable "favorable" resources,to design a channel-adaptive space-collaborative constellation and to propose a fast maximum likelihood detection algorithm.1.Channel-adaptive space-collaborative constellation design.As for the MIMO-VLC with low correlation channel,a universal channel-adaptive space-constellation design criterion is established.For this criterion,in one way,the optimization model of constellation considers the channel effect,and the objective function is based on the minimum distance of the receiver constellation.In another way,different from the current techniques,we establish the relation between the LEDs,which transforms the uncontrollable ICI into the controllable resources and thus improves the error performance.Based on this criterion,by integrating the mathematical knowledge of spatial geometry and ensemble division,we design a regular CASCC in terms of the base station and the high-order constellation for MIMO-VLC channel.Simulations show that our CASCC not only has better error performance,but also has better adaptability to the channel,compared with the conventional scheme for MIMO-VLC.2.Fast maximum likelihood detection algorithm.For the CASCC,the constellations at the transmitter and that at the receiver is one-to-one.Moreover,the entire constellation at the receiver can be represented by two base vectors.It is both regular and known.Based on these characteristics,a fast maximum likelihood detection algorithm is designed by reducing the maximum likelihood search area.We compare it with the conventional maximum likelihood detection from the perspective of the theoretical analysis and simulation.The results show that our fast detection algorithm reduces the computational complexity,without the loss of error performance.