Study On The Related Parameters Of High-Speed Mobile Channel

With the rapid development of the communications, the urgent issue to be resolved is toimprove the utilization rate of spectrum. However, MIMO is the key technology to solve thisproblem, which can greatly improve the utilization rate of spectrum and channel capacity. OFDMtechnology is based on the multi-carrier transmission. Under the condition of frequency-selectivefading channel, since the simple MIMO system cannot ensure the good characteristics mentionedabove, so its frequency-selective fading channel can be divided into several frequency-non-selectivefading channels on the sub-carriers by using OFDM, and the combination can realize theadvantages of MIMO. This paper can be summarized as follows:First, the source and significance of this study are given by analyzing the current status ofresearch according to the development trends of the mobile communication technology. Based onthe characteristics of the wireless channel, the basic principles and system models of MIMO,OFDM, and MIMO-OFDM system are introduced respectively.Then, The channel estimation under the frequency-non-selective and frequency-selectivechannel environments are mainly researched for MIMO system. First of all, the theoreticalfoundation of channel estimation is introduced briefly and the observation results are analyzedrespectively by using LS, ML and MMSE algorithm. Then, the algorithm estimation under differentchannel environment is derived according to the applied training sequence, and the correspondingsimulation results are given at the same time.Finally, this paper studies the algorithm of carrier frequency offset estimation forMIMO-OFDM system which needs perfect estimation, because the loss of the orthogonality amongsub-carriers can be caused by carrier frequency offset, which will reduce the system performance.As for the system with time-domain sequence, two sequences with different length are used toestimate algorithm based on the different phase offsets of the same training sequence reflected intime domain from frequency offset. As to the frequency domain, the method of frequency-domainpilot is employed, in which the algorithm estimation is carried out by relying on the detection ofadjacent pilot phase. And the corresponding simulation results of different algorithms are presented.