Hybrid Precoding Technology Research In Millimeter Wave Massive MIMO Systems

With the development of wireless communications,the current radio band is becoming more and more crowded,extending communication radio frequency(RF)into millimeter wave(mm Wave)band seems the essential one for the severe spectrum gridlock.Nevertheless,mm Wave signals easily suffer from severe path fading during propagation,so the conventional small-scale transceiver architecture would hardly allow the receiver to obtain sufficient received signal power.Therefore,massive multi-input and multi-output(MIMO)antenna array,the most promising transceiver structure,has been proposed for the emerging 5G/6G communication systems as its large spatial degree-of-freedom enables to greatly enhance signal gain via transmitter precoding and receiver combing.In massive MIMO systems,the full-digital precoding requires a dedicated RF chain and the corresponding matching circuit for each antenna.It is difficult for massive MIMO systems to suffer such high power consumption and hardware costs.The hybrid precoding/combing is an implemented technique to reduce the number of RF chains while achieve almost the same performance with full-digital precoding in massive MIMO systems,which has received extensive attention from researchers.The completed work is summarized as follows.1.Based on the singular value decomposition(SVD)technique and the concept of equivalent channel,joint hybrid precoding strategies with low complexity are proposed to maximize spectral efficiency for single-user massive MIMO systems.Specifically,the algorithm decouples the design of the analog precoder/combiner from the digital precoder/combiner,after designing the analog precoder and the analog combiner,the corresponding equivalent channel is constructed.Then,the digital precoding and combining are realized directly by applying the SVD technique to the matrix of equivalent channel.The simulated results show that our proposed schemes can achieve superior performance with lower complexity to the existing ones.2.The scenario of massive MIMO systems is extended from single-user to multi user,and a multi user hybrid precoding algorithm based on zero-forcing(ZF)algorithm is designed.In details,each user constructs the corresponding equivalent channel according to the single-user hybrid precoding algorithm as previously described.Then,the total equivalent channel of all users are obtained.Finally,the ZF algorithm is used to eliminate the interference between users.The simulation results show that the designed multi user hybrid precoding algorithm is close to the performance of the full-digital ZF precoding algorithm.3.For mm Wave massive MIMO systems,a regression model based on the support vector machine technique is proposed,through which the optimal number of data stream can be obtained automatically based on the propagation characteristics of wireless channel,the number of antennas,and RF chains to maximize the system capacity.Simulation results show that the performance of hybrid precoding are effectively improved after using the proposed model.