Subspace-based Blind Channel Estimation In MIMO-OFDM System

The premise of the studying of a wireless communication system is to know thewireless channel. Only the transmitter and receiver is designed by knowing the channelinformation, the optimal performance of system can be obtained. Therefore, the channelestimation occupies a pivotal position in the system design. The channel estimationmethod is mainly divided into two categories: non-blind channel estimation based onpilot and blind channel estimation method. The former method should send pilotcontinuely when sending signals, which greatly reduces the bandwidth efficiency of thesystem. For MIMO-OFDM system which is used widely, the efficiency of the method islower and the system is more complex because of the more number of antennas inMIMO-OFDM system. Thus the demand for high spectral efficiency spurred extensiveresearch on the blind channel estimation method. The blind estimation becomes acurrent research focus. This paper mainly researchs the blind channel estimation methodfor MIMO-OFDM system. And the matlab simulation is used to verify the performanceof the algorithm which is provided by this paper.Firstly, the theoretical basis of the propagation mechanism of the wirelesschannel, the declin of the radio channel, and the basic channel model areresearched in this paper. The basic principles of MIMO-OFDM system isconducted in depth learning. The basic principles of OFDM technique are anfocus which include its system structure, modulation and demodulation method,and OFDM guard interval. The MIMO channel capacity under differentconditions is analyzed. A system model is provided by combining MIMO andOFDM.Secondly, the mathematical model of the SIMO system channel estimation isanalyzed. Based on the model, the second-order statisitics-based blind channelestimation algorithm is studied in detail. A mathematical derivation of the entirealgorithm is provided, and the performance by NRMSE is simulated. Then on thebasis of the existing literature, an improved MIMO-OFDM system model isproposed. And the subspace-based algorithm is used in the improved model toestimate the channel information. The ambiguity matrix in the MIMO system isanalyzed, and the solution is provided. Finally, the theory of oblique projection is researched. And the obliqueprojection subspace method is applied in the model that is proposed in this paperto estimate channel information. And the algorithm is improved and adjusted tothe model. Finally, the matlab simulation is used to verify the performance of thealgorithm.