Low-Rank Channel Estimation for MIMO MB-OFDM UWB System Over Spatially Correlated Channel

In this letter, we derive a low-complexity channel estimation scheme for multiple-input multiple-output multiband orthogonal frequency division multiplexing ultrawideband (MIMO MB-OFDM UWB) system over spatially correlated channel. More precisely, we drive a linear minimum mean-squared-error (LMMSE) channel estimation algorithm. Then, an optimal low-rank LMMSE channel estimation method based on low-rank approximation is deduced from the LMMSE one. Through simulation results, we show that the proposed LMMSE channel estimator provides significant performance gain over least square (LS) one. We also show that the low-rank LMMSE channel estimator reduces significantly the computational complexity while maintaining comparable performance with LMMSE one for low signal-to-noise ratio (SNR) and minor performance degradation for high SNR.