A Novel Subspace Tracking Based Blind Channel Estimation of MIMO-OFDM

Subspace methods have been widely used in blind channel estimation of MIMO-OFDM system. However, direct subspace decomposition suffers from heavy computation complexity of o(N)³. In this paper, a new data projection method (DPM) based principal and minor subspace tracking algorithm is presented by exploiting a novel orthonormalization matrix. The new algorithm exhibits good numerical stability and its dominant complexity reaches the 3NL lower bound. Simulation results show that the proposed channel estimation scheme has better performance compared with the original one.