Research On Channel Estimation Algorithm Of LTE System And FPGA Implementation

As the development direction of next generation mobile technology, LTE (Long Term Evolution) can provide users with high-speed rate and high broadband. It is an communication technology standard to solve the users’ demand for mobile broadband. As is known to all, multiple input multiple output (MIMO) and orthogonal frequency division multiplexing (OFDM) technology is the core technology of LTE system. They can effectively improve the data rate and increase the system capacity. With the rapid development of high-speed railway construction in our country, the terminal equipment of the high speed makes communication channel become more complicated and brings new challenges to the wireless communication system. As for MIMO-OFDM system, how to obtain accurate channel state information is the basic conditions to realize the MIMO-OFDM coherent transmission which directly determine the error probability of the receiver’s MIMO detection. In addition, transmitter’s precoding needs to obtain accurate channel information as well. Therefore, the thesis will focus on analysis and study of channel estimation technology of LTE system based on MIMO-OFDM.First of all, the thesis introduces the basic principle of MIMO and OFDM, the two key technologies of LTE system, and wireless channel characteristics, that illustrates the difference between slowly varying channel and fast varying channel and presents the wireless channel simulation model.Then, according to the LTE physical layer protocol, the thesis sets up the down-link LTE system. On the basis of the system, the thesis makes a Comparison and analysis of several current mainstream’s channel estimation methods, including LS, LMMSE and SVD-LMMSE algorithm, and studies the data interpolation algorithm, including linear interpolation algorithm, the second order polynomial interpolation algorithm, cubic spline interpolation algorithm and interpolation algorithm based on DFT, whose performance simulation analysis is given. In order to adapt to bigger Doppler frequency shift, the thesis studies channel estimation method based on basis expansion model (BEM) which is suitable for time-varying channel. Firstly, the thesis introduces the basic principle of BEM model and several common BEM model, including CE-BEM, P-BEM, DKL-BEM and DPS-BEM, discusses and simulates the performance affection of different channel BEM algorithm in different the Doppler frequency shift and BEM order. Secondly, the thesis analyzes the principle of channel estimation algorithm based on BEM, and gives the simulation analysis of CE-BEM and DPS-BEM, the two kinds of channel estimation based on BEM model.Finally, combined with the previous algorithm’s principle analysis, the thesis gives the FPGA hardware implementation diagram and the Moldesim simulation results of the LS channel estimation algorithm in the slowly varying channel, by the way, compares the algorithm’s performance of FPGA with MATLAB’s.