Semi-Supervised Receivers for MIMO Systems with Multiple Khatri-Rao Coding

In the last years, tensor based semi-supervised receivers for Multiple-Input Multiple-Output (MIMO) point-to-point wireless communication systems have been extensively used. In this paper, we propose the Multiple Symbol matrices Khatri-Rao (MSMKR) coding for a point-to-point MIMO system and two tensor based semi-supervised receivers. The tensor of received signals is modeled as a PARAllel FACtor (PARAFAC) model. Tensor based semi-supervised iterative and non-iterative receivers are proposed to simultaneously estimate symbol and channel matrices. Simulation results are provided to evaluate the performance of the proposed receivers in terms of symbols and channel estimation.