DFT-Based Channel Estimation Performance Analysis and a Robust Estimation Algorithm in MIMO OFDM Systems with Virtual Carriers

MIMO-OFDM systems can effectively improve system capacity without increasing channel bandwidth and transmit power. However, they need more accurate channel estimation. Conventional DFT-based channel estimation (CE) suffers from the dispersive distortion due to virtual carriers (VCs). In this paper, the performance of time domain filtering least square (LS) CE algorithm for MIMO-OFDM systems with VCs is analyzed. A concise expression which reveals the relationship between the maximal estimation error variance among all the carriers and the design parameters is provided. Then a robust CE method is presented. Simulation results show that this method can improve estimation performance significantly even when the channel length information is overestimated