Blind identification of MIMO channels: a closed form solution based on second order statistics
Victor A. N. Barroso, João Xavier · 2003
We address the blind multiple-input multiple-output (MIMO) digital channel identification problem. Our approach is based on the source correlative coding concept. Spectral asymmetry is induced, permitting the receiver to recover the MIMO channel from the observed data, by a second-order statistics (SOS) matching strategy. No additional power or bandwidth, nor synchronization between the sources, are required, and the original data rate is maintained. We present a new identifiability theorem which guarantees uniqueness of the MIMO channel estimate (up to a phase factor per user). A closed form algorithm attaining that identifiability bound is introduced. Computer simulations are presented. When compared with the transmitter induced conjugate cyclostationarity (TICC) approach, the results obtained show that our solution provides lower probability of error and a superior robustness to carrier frequency misadjustments.