Research On Key Technologies Of Large Scale MIMO Transmission

The large multiple input multiple output(MIMO)technology is one of the most important technologies in the 5G communication.In order to optimize the transmission performance of the large MIMO system,this paper did some research on the signal processing technologies in the large MIMO system at the transmitter side and the receiver side respectively.In order to alleviate the high computational complexity of the detection algorithm in the large scale MIMO system,this paper proposed two novel low complexity detection algorithms.The improved hybrid Gibbs sampling detection algorithm reduces the computational complexity by adding the euclidean distance threshold decision which terminates the iteration in advance.Additional,it optimizes the disturbance which is added in the measured signal distribution to improve the accuracy of detection and avoid iteration non-convergence.On the other hand,the improved ordered successive interference canceling interference algorithm is based on the discovery that the signals in the intersecting area of the circles are less reliable,while the centers and radii of the circles are the constellation point of the modulation constellation and the power of the interference respectively.Based on the above discovery,this paper modified the criterion of judging signal reliability to improve the bit error rate performance and offsetted the interference in group to achieve the trade-off between the computational complexity and the bit error rate performance.With frequency selective fading channel,large MIMO orthogonal frequency division multiplexing(OFDM)syetem employes the hybrid analogue and digital beamforming algorithm to reduce the number of the radio frequency(RF)chains.However,the existing hybrid analogue and digital beamforming algorithms need scheduling beteween the users and subcarriers which sacrifices the sum rate performance of the system.To address the problem,this paper proposed a novel hybrid analogue and digital beamforming algorithm without scheduling in the large MIMO-OFDM system with frequency selective fading channel.This algorithm reduces the number of the RF chains in frequency selective fading channel to the same as that of the flating fading channel.Besides,its computational complexity increases linearly with the number of transmit antennas.Finally,the simulation results show that the proposed beamforming algorithm provides a sum rate performance which is close to the full digital beamforming algorithm,and this algorithm can support multi subcarriers to improve the sum rate performance of the system because this algorithm does not need scheduling between the users and subcarriers.In the spatial modulation(SM)-OFDM system,this paper analysed the analytical BER upper bound and derived the optimal power allocation between the data and the pilot symbols by minimizing the upper bound with minimum mean square error(MMSE)channel estimation when employing all generalized linear interpolation techniques.Simulation results show that the proposed optimal power allocation scheme provides a substantial gain in the average BER performance for the SM-OFDM system compared with equal power allocation.To achieve sidelobe suppression,the SM-OFDM system is combined with N-continuous(NC)filter.This paper proposed an approximate system model for the SM-OFDM system with NC filter based on the investigation on the statistical characteristics of the interference introduced by the NC filter.Based on the proposed approximate system model,the average BER performance of the system operating with perfect channel estimation and MMSE channel estimation is analysed separately.Furthermore,this paper derived the optimal power allocation between the data and the pilot symbols by minimizing the upper bound.Finally,the numerical results show that the NC filter can degrade the BER performance of the SM-OFDM system,and the analytical BER closely matches its simulated counterpart under different parameter configurations.Compared with equal power allocation,the proposed optimal power allocation achives a substantial gain in the average BER performance of the SM-OFDM system with NC filter.Finnally,this paper summarized the research results of the full text and indicated the following research direction.