Research On Single Carrier Based Millimeter Wave MIMO System

Due to traditional sub-6GHz communication technology not meeting the growing mobile data,high-frequency millimeter wave communication has become a key technology to solve this problem,which has a large number of available bandwidth resources.Although millimeter wave communication has been standardized in WLAN and WPAN,but there are still many challenges in the wide use of millimeter wave communication in mobile communication system.First of all,due to the short wavelength,millimeter wave has serious propagation loss in space,which shortens the communication distance.Secondly,due to the wide bandwidth and high frequency band,millimeter wave communication has more complex baseband signal processing and higher the hardware cost of RF chains.Moreover,the use of large-scale antenna array at base station further increases the hardware cost and power consumption.Although OFDM can effectively tackle ISI in broadband communication,the use of CP structure still decreases the spectrum efficiency.Besides,OFDM is sensitive to CFO and has higher PAPR,which brings down its application in specific scenarios.Due to the weak coverage of millimeter wave,it is particularly important to improve the efficiency of the power amplifier without increasing the additional transmission power.Compared with the multi carrier waveform represented by OFDM,single carrier has lower PAPR and higher power amplifier efficiency.Thus,single-carrier based millimeter wave communication system is studyed,including SC-FDP system on massive MIMO and hybrid precoding of single carrier millimeter wave based on lens antenna array.The main research and contributions are as follows:First,the waveform principles are introduced,including single carrier waveforms,OFDM and new multi-carrier waveforms.Moreover,the advantages and disadvantages of various waveforms are analyzed based on simulation results.Compared with the multi-carrier waveforms,single carrier has lower PAPR and lower complexity,which means that the single carrier wave can make up for the shortcomings of large-scale millimeter wave MIMO,providing a theoretical basis for subsequent research.Second,a joint frequency domain equalization and precoding design based on single carrier is proposed for frequency selective channel in large-scale MIMO downlink communication system.Specifically,the frequency selective channels are transformed into flat fading channels by using single carrier frequency domain pre equalization at the base station,and the expression of SINR in flat fading channels is obtained.Based on the obtained SINR expression,the traditional ZF and MF precoding are used at the base station to eliminate inter-user interference and inter-block interference.In addition,the asymptotic expressions of ZF and MF precoding matrices and the reachable sum rate under the precoding scheme are derived by using the orthogonality of large-scale MIMO channels and the theory of large-scale random matrix.Finally,the approximate ZF and MF are used to calculate the PAPR of SC-FDP and OFDM respectively.The simulation results show that SC-FDP has similar system performance compared with OFDM system.Besides SC-FDP has lower PAPR than OFDM,while the base station doing almost all computing.At last,a single carrier transmission scheme based on three-dimensional lens antenna array is proposed for millimeter wave broadband communication.Due to the fact that the lens antenna array can effectively separate the signals from different directions and the sparsity of millimeter wave channel,the broadband signal is transformed into multiple parallel narrow-band signals by time-delay compensation at the base station.Then,the SINR expression of each user is derived.Finally,the optimal transmission and reception beamforming is designed based on the system sum-rate maximization criteria.Simulation results show that the single carrier transmission scheme based on lens antenna array has comparable system performance as MIMO-OFDM based on planar array,while the former has lower system complexity and power consumption.