Study Of The Pilot Matrix Designing In Channel Estimation Of Massive MIMO System

Massive multi-input multi-output technology can effectively improve the performance of the wireless communication system,and the characteristics of the channel state information have a vital impact on the system performance.Channel estimation using the pilot signals to obtain channel stat information is one of the common channel estimation methods,where a certain number of pilot sequences are added to the transmit signal,and the receiver obtains the channel parameters of the pilot point and the data point respectively through the appropriate processing technology,which is the most widely used method.Therefore,how to design the pilot matrix with excellent performance is the first key content of this thesis.The channel matrix of Massive MIMO system channel has sparse characteristics,so channel estimation can be performed using compressed sensing theory,that is,channel estimation by compressed sensing reconstruction algorithm;the pilot optimization criterion of compressed sensing channel estimation algorithm is the guarantee whether the sensing matrix can achieve good reconstruction effect.How to design a good perception matrix based on the channel estimation based on the OMP algorithm of compressed perception theory is the second key content of this thesis.In conclusion,two pilot optimization schemes are studied respectively:(1)Pilot frequency optimization design method based on the is angular compact framework theory with low complexity.Theoretical lower bounds on the column correlation of the pilot matrix during channel estimation in Massive MIMO systems are obtained using the is angular compact framework theory;and the CSM(Cholesky and Sherman-Morrison)iterative algorithm is adopted in the channel estimation algorithm to reduce the computational complexity of the channel estimation.Simulation results show that the proposed algorithm can obtain a pilot matrix with lower auto correlation than the previous algorithm.(2)The pilot optimization design method based on the minimum channel reconstruction error rate under the framework of compression-sensing OMP algorithm,and the reliability of the proposed algorithm is proved theoretically,and extended to a more general signal reconstruction algorithm framework.First,under the framework of OMP algorithm,with the goal of channel reconstruction error rate minimization,the relationship between pilot matrix auto correlation and channel reconstruction error rate is analyzed theoretically analyzed,and then derive the two-point criteria to follow for pilot matrix optimization: pilot matrix column correlation expectation minimization and variance minimization.Then the pilot matrix optimization criterion is further extended to the framework of a general sparse recovery algorithm,making the optimization criterion universal.According to the optimization criteria,the adaptive correlation matrix reduction parameter pilot optimization algorithm is designed.The simulation results show that the proposed pilot matrix has better matrix and column correlation properties compared with the random Gaussian matrix,Elad method and a lower channel reconstruction error rate.