The technology of wireless communication has extraordinary practical value as well as special difficulties. The combination of MIMO and OFDM has greatly improved the communication capacity and transmission rate, in the meantime, MIMO-OFDM system can effectively combat multi-path fading and interference. So this system has been considered as one of the most available scheme in beyond 3G/4G wireless communication system. In order to improve system performance in application, channel estimation technique plays a quite meaningful and challengeable role in MIMO-OFDM system. This paper has studied the technique of channel estimation in MIMO-OFDM system by both theoretical analysis and computer simulation. The main jobs in this paper is as below four: 1. Deeply analyzed morden technique of MIMO-OFDM system in both domestic area and abroad, summarized the two main trends of those study, namely, space multiplexing based on OFDM and STC-OFDM. After studied the principles of two systems respectively, the structure character of STTC-OFDM system is discussed in detail, also, the chatarter of MIMO fading channels is introdueced. In the end, the importance of channel estimation in MIMO-OFDM system is pointed out. 2. Deeply studied the channel eatimation in MIMO-OFDM system based on training sequence. Analyzed the LS algorithms in time domain and in frequence domain. Mordified algorithms has been proposed to improve the system performance, at last, computer simulation has been conduct, testifying the vadility and feasibility of these algorithms. 3. Completely stuied the pilot-based channel eatimation in MIMO-OFDM system. The criterias of pilot selection and distribution are analyzed first. After that, comb-type pilot based and two-dimensional pilot based estimation algorithms are studied. Modified methods have been proposed. At last, the performance comparison of those algoritghms is conduct after computer simulation. 4. Aimming at fast fading time varying MIMO channel, two channel estimation algorithms based on OFDM CP is proposed. One is based on implicit training sequences, the other is based on CP reconstruction. Two algorithms both can increase the transmission rate and frequency efficiency. The completely analysis and comparison of these algorithms are made by computer simulation, by which the vadility of two methods are proved as well. |