| Visible Light Communication(VLC),as a new type of communication method,has rich spectrum resources,extremely high transmission rate,and is not susceptible to wireless electromagnetic interference.It has received extensive attention and research in recent years.Massive Multiple-Input Multiple-Output(MIMO)technology has become one of the core technologies of the 5th generation mobile communication.Applying massive MIMO technology to a visible light communication system can improve system performance and coverage by merely adding a spatial link without increasing transmission power and bandwidth.At present,most researches on visible light communication are concentrated in indoor scenes.However,in outdoor visible light communication scenarios,the channel environment is mainly affected by factors such as atmospheric turbulence and atmospheric attenuation,resulting in complex and variable channel environment,which seriously affects outdoor visible light communication performance.Based on this,this paper mainly studies the semi-blind detection and modulation constellation optimization algorithm of outdoor visible light communication system.Firstly,the channel characteristics of outdoor visible light communication are studied,and the corresponding channel models under different atmospheric turbulence conditions are summarized.Based on the non-negative characteristics of optical signals,a maximum likelihood detection algorithm based on Gaussian approximation for arbitrary channel models is proposed.When the transmitting and receiving antennas are large,the algorithm determines that the receiving signal approximates the Gaussian distribution according to the central limit theorem and the large number theorem,and then obtains the asymptotically optimal decision threshold of the received signal with arbitrary modulation order,in order to minimize the system average symbol error rate.According to the main component of the system average symbolic rate: exponential decay term,the modulation constellation is optimized by using the criterion of maximizing the minimum exponential attenuation coefficient,and the numerically iterative solution method is used to obtain the optimized modulation constellation.Secondly,a maximum likelihood detection algorithm based on Laplace approximation is proposed for the channel environment obeying the negative exponential distribution model.Based on the analysis of the probability density function of the channel model,the algorithm approximates the known channel model by Taylor's formula expansion,and obtains a more accurate judgment threshold expression,and derives the average error symbol rate expression of the system performance.The simulation results show that the average symbol error rate performance of the algorithm is significantly better than that of the Gaussian approximation based detection algorithm when the transmit antenna is large.Finally,considering the optimal decision threshold of the system,we pay more attention to the fitting of the probability density function of the approximated channel model and the original channel model at its truncation.Therefore,a segmentationbased Gaussian approximation is proposed.Maximum likelihood detection algorithm.Under the negative exponential distribution model of the channel environment,the probability distribution of each sub-channel superimposed has a positive skew characteristic,which approximates the received signal into two Gaussian distributions,and then obtains a more accurate optimal decision threshold.The expression is closed,and according to the average error rate expression of the system,the optimized modulation constellation is obtained by numerical iterative solution under the condition of average optical power constraint.The simulation results show that compared with the other two proposed algorithms,the algorithm can reduce the average symbol error rate of the system in the large-scale case of the transmitting and receiving antennas. |
PDF Full Text Request