Research And Implementation Of Massive MIMO Precoding Technology

MIMO(Multiple-Input Multiple-Output)is a key technology in LTE-A(Long Term Evolution Advanced)which is required to use multiple antennas at the transmitter and receiver respectively,and the signal is transmitted and received by multiple antennas.MIMO can improve communication quality.Without increase of spectrum resources and transmitting power,MIMO technology can make full use of the space resources and increase the system capacity.And as a key technology in MIMO system,precoding technology plays an important role in eliminating the interference between channels and improving the spectrum efficiency.However,with the application and popularity of massive terminal equipments and new services,the demand for the wireless communication transmission rate is more and more high.It is estimated that by 2020,the demand for wireless communication transmission rate will be 500-1000 times the LTE-A system,but the current LTE-A system will not be able to meet the requirements of the transmission rate.Therefore,in order to improve the system capacity,the future 5G mobile communication will use Massive MIMO(Massive Multiple-Input Multiple-Output)technology that is equipped with a large number of antennas(more than 100)in the base station and provides services for multiple single antenna terminals.However,with the increase of the number of base station antennas,even simple linear precoding(e.g.ZF precoding,RZF precoding)has a high computational complexity.Therefore,how to reduce the computational complexity of precoding becomes the focus of this paper.Firstly,the system model is introuced and the spectrum efficiency of this system is analysed in this paper.Secondly,the current popular three linear precoding algorithm are studied,namely ZF(zero-forcing),RZF(regularized zero-forcing)precoding,MRT(Maximum ratio transmission)precoding,and the bit error rate,spectrum efficiency and complexity of three linear precoding algorithm are simulated and analyed.And then the low complexity TPE(truncated polynomial expansion)and MDPE(matrix decomposition-polynomial expansion)precoding algorithm are studied that can reduce the computing complexity by taking the first L items from polynomial expansion of matrix approximation to replace the matrix inversion in the RZF precoding algorithm.In this paper,the bit error rate,spectrum efficiency and complexity of TPE and MDPE algorithms are simulated and analyzed,and the specific scenarios of the two algorithms are discussed.In addition,the method of calculating the coefficients of MDPE precoding algorithm is improved in this paper.The simulation result shows that the performance of the algorithm is improved.Finally,the module encapsulation of ZF encoding algorithm by GNU Radio software platform is realized,the algorithm is verified in 8×8 MIMO system.