ADAPTIVE CHANNEL ESTIMATION FOR MIMO-OFDM SYSTEM IN MULTIPATH FADING CHANNEL

Abstract

Channel Estimation is a critical task in wireless communication. It is one of the vital parts of the wireless communication which is a method used to significantly improve the performance of the system. OFDM can be associated with number of antennas at the sender and receiver sides called MIMO-OFDM system. In this thesis, Adaptive Channel Estimation techniques such as Least Mean Square, Recursive Least Square and Kalman filtering are used for the MIMO-OFDM system in the Rayleigh and Rician fading channels. The various diversity configurations like 2x1, 2x2, 2x3 and 2x4 for the MIMO-OFDM have been used. Performance of the high order diversity is better than the low order diversity system i.e 2x4 has the best result in terms of BER. Simulation results show that RLS has better performance than LMS. The performance of the Kalman filter is better than the LMS and RLS techniques. The effect of the Doppler shifts has been studied for the different MIMO system to analyze the time varying environment of the system. The throughput analysis has been done. Performances are measured in terms of the bit error rate vs SNR and Throughtput vs SNR. While analyzing computational complexity of the LMS, RLS and Kalman filter, LMS algorithm is less complex than RLS and kalman filter algorithm. Simulation has been done to verify the proposed method.