SPC04-6: Channel Estimation for MIMO-OFDM Systems in Rapid Time Variant Environments Based On Channel Statistics Estimation

This paper examines pilot-aided channel estimation algorithm for multiple-input, multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems in rapid time-variant environments. It includes Doppler frequency estimation, delay path profile estimation and interpolation algorithms in time and frequency domains. The algorithm performs channel statistics estimation for each frame and uses the pilot symbols of this frame where all signals pass through the same propagation channel. The proposed algorithm uses separable, interpolation order changing to reduce channel estimation complexity. This algorithm first applies interpolation in the time domain and then in the frequency domain. Hence, for the pilot pattern of IEEE 802.16 e we get twice as many reference symbols for frequency interpolation, which improves the interpolation quality, especially for channels with long delay profiles. Computer simulation of the V-BLAST structure demonstrates high effectiveness of this algorithm and its robustness to increasing subscriber speed. Packet error rate (PER) degradation, in comparison with the ideal channel, is l-2.5 dB for 4 x 4 mode and 0.5-1.5 dB for 2 x 2 mode. Our results show that the performance degradation is acceptable not only up to 250 km/h, but also up to 500 km/h.