Design And Analysis On Channel Estimation Technique In MIMO-OFDM System

With the rapid development of wireless communication technology,people’s demand for video,graphics,network and data is growing fast, a wireless communication system with large-capacity,high-speed,multi-service and high-reliability will be the trend of future. MIMO-OFDM fits the needs of next generation communication systems,becase it can not only improve channel capacity but also resist multipath fading effectively,and it has a very broad prospects of development.Channel estimation directly decides the communication quality of MIMO-OFDM,so channel estimation techniques of wireless communication systems are mainly researched on this thesis.In this thesis,MIMO,OFDM and MIMO-OFDM system are studied.the training sequence based channel estimation and the pilot symbol based channel estimation are discussed deeply.For training sequence based channel estimation,comparison of the PN sequence based LS channel estimation and optimal training sequence based LS channel estimation show that the PN sequence can not only overcome the disadvantage of LS algorithm which caused by the matrix inversion,but also can improve the accuracy of estimates.Because of these advantages of PN sequences, this thesis attempts to combine PN sequences with the decision feedback algorithm to improve the initial estimation accuracy of the decision feedback algorithm. Theoretical analysis and matlab simulation results show that the new proposed algorithm can indeed improve the accuracy of the initial estimate accuracy. For pilot based channel estimation, this thesis comparative analysis the interpolation algorithms both in time domain and transform domain. In time domain an improved DFT interpolation algorithms is proposed to reduce the noise component in the CP.And in transform domain,an improved DCT interpolation algorithms is put forword through combining it with the decision feedback to alleviate energy leakage and reduce the channel noise. Simulation results show that the two improved algorithms are better than their base algorithm in suppressing the channel noise.