Channel Estimation Algorithm And Its DSP Implementation For LTE Uplink In High Mobility Environments

With the increasing needs in railway service, people’s demands for transport capacity and transmission rate of communication is are also improved. LTE, as one advanced communication system using a combination of OFDM and MIMO technology, has the merits of high data rate, low system time delay, large system capacity, and wide coverage. It is able to meet these newly emerged requirement for railway communication. So, it is particularly urgent to study the LTE channel estimation algorithm, enabling correct signal detection in receiver in high mobility environment, especially for high-speed railway scenario, where the user speed is relatively high and the doppler frequency shift is usually larger, to accurately estimate the channel coefficient more important. Therefore this thesis mainly studies the channel estimation algorithm under LTE standard, and under the basis of theoretical analysis, choose a algorithm with lower complexity but excellent performance under high mobility environment and then implement it in DSP hardware.In order to select the optimum channel estimation algorithm, this thesis firstly describes the reception model in frequency domain of MIMO-OFDM under basis expansion model. And then according to the special structure of LTE uplink pilot, the traditional LS channel estimation algorithm, DFT algorithm, MMSE algorithm and the BEM based maximum likelihood algorithm and Kalman filter algorithm are analyzed. With the consideration of ideal synchronization at the receiver, the thesis has done simulation analysis and complexity evaluation for above several algorithms in different channel scenarios. The simulation results show that in high mobility environment ML based BEM model can be used for DSP algorithm implementation.This thesis adopts TMS320C6474 DSP hardware platform, and proposes the DSP implementation solution for the BEM based ML algorithm. Meantime, it realizes the algorithm design and optimization of each module in software integrated development environment CCS. The error analysis of the algorithm implementation of each module is also given. Finally, the channel estimation algorithm modules are integrated for debugging and then the selected channel algorithm is verified in the LTE link implemented with DSP.