Semi-blind channel estimation in MIMO commumication by tensor factorization

In this paper, we introduce a tensor-factorization method for solving channel estimation problem in MIMO applications. we address the estimation problem through a 3-way tensor analysis. We represent the received signals in a 4times4 MIMO system as a third-order tensor with modes: receiver antennas, packet symbols, and number of packets. Then we demonstrate that the multi-way analysis of PARAFAC2 can be successfully used for solving the MIMO channel estimation problem. This method uses different m-Sequence training symbols for different transmitted signals on the transmitter side. For evaluating the method we applied the well known MMSE-VBLAST detection method to the estimated channel and the results are reported.