Research On MIMO-OFDM Transmission Technologies In High Speed Optical Wireless Communications

To meet the the increasing demand of wireless communication,multi-input multiple-output(MIMO)technology with large channel capacity and spectrum efficiency,as well as orthogonal frequency division multiplexing(OFDM)technology with anti-multipath fading are adopted.However,in optical wireless communications,there are still some problems such as unstable free space optical channel transmission,high detection complexity,high transmission error rate,and so on.In order to solve the above problems,the key technologies of optical wireless channel modeling and signal processing based on MIMO-OFDM system are mainly studied in this paper.Aiming at the shortcomings in existing optical wireless channel model,OFDM modulation and signal detection,a minimum mean square error sequential interference cancellation(MMSE-OSIC)signal detection strategy for asymmetric limiting optical-orthogonal frequency division multiplexing(ACO-OFDM)modulation is proposed.The optical channel model of an optical wireless system,the mean square error of the ACO-OFDM channel estimation,the error detection performance and computational complexity of the MMSE-OSIC algorithm are studied in detail.The contents and contributions of the thesis are concluded as follows:1.According to the problem that the analysis of optical wireless channel is difficult due to turbulence and light scattering in free-space optical wireless communications,the modeling of optical wireless channel is studied.First,on the basis of the original weak turbulent channel,the alignment error factor is introduced,and the causes leading to alignment error in practice are summarized and qualitatively analyzed.Then,the quantitative optical channel model with the alignment error is established.The outage probabilities of the system under different alignment errors are derived and the theoretical expressions of channel diversity gain is calculated.Finally,the asymptotic expression is obtained under the condition of large signal-to-noise ratio.The simulation results show that the optical channel with alignment error is closer to the real optical signal transmission condition,and the diversity gain of the system is only related to the alignment error value.When there is no alignment error,the diversity gain is directly proportional to the number of transmit and receive antennas.2.According to the difficulties of the optical signal modulation and demodulation and channel estimation in optical wireless MIMO-OFDM system,an ACO-OFDM channel estimation scheme based on superimposed periodic training sequence is proposed.First,in order to enable OFDM technology to be directly used in optical wireless communication,asymmetric limiting and Hermitian symmetry techniques are used to make the OFDM signals unipolar.Second,to solve the problem that optical wireless transmission is sensitive to noises and the mean value of signals is non-zero,a channel estimation based on the superposition training sequence and a linear least mean square error criterion are proposed.By improving the training sequences and local matrix,the estimation performance and transmission efficiency are effectively improved.Finally,combined with discrete fourier transform precoding and partial transmission sequence(PTS)algorithm,the peak-to-average ratio(PAPR)suppression is obtained and verified by numerical simulations.3.According to the problems of error propagation and high computational complexity in signal detection of optical wireless MIMO-OFDM systems,a feedback-information-assisted sequential interference cancellation(OSIC)algorithm is proposed,which combines iterative approximation and multi-layer joint detection strategy together.In this scheme,soft estimation of each layer is used as feedback information,and feedback candidate points are selected from constellation points to construct candidate vectors.Then,Maximum likelihood(ML)criterion is used to select the vector closest to the transmitted signals from the set of candidate vectors as output,so as to reduce the error caused by early iterative detection.In addition,the iterative approximation strategy with quantitative relaxation factor and initial iteration value is used to simplify the matrix inversion process involved in minimum mean square error(MMSE)criterion.Finally,the multi-layer joint detection is used to optimize the OSIC algorithm,which can save some iteration detections.Simulation results show that the proposed algorithm can improve the detection performance by 4-5 dB with equal error rate,and the computational complexity decreases from the cubic power to the quadratic power.Through the optimization of channel modeling,channel estimation and signal detection are investigated in optical wireless MIMO-OFDM systems,the BER performance is effectively improved and the complexity is reduced.Therefore,the proposed scheme can be widely and efficiently applied in the research of optical wireless communication systems.