Error Regularized Turbo-MIMO MMSE-SIC Detection with Imperfect Channel State Information

In our previous study [1], we have found that the performance of low-complexity successive interference cancellation (SIC) detection algorithms in Turbo multiple-input multipleoutput (MIMO) systems can be improved significantly by a simple regularization technique that mitigates the propagation of detection errors during the detection and the decoding cycle. Since SIC based detection is prone to imperfect channel state information (ICSI), in this work we extend our study to take into account the effect of ICSI resulting from least squares (LS) Gaussian channel estimation aided by orthogonal pilot sequences. We introduce the ICSI-aware linear, SIC, and maximum-likelihood (ML) Turbo- MIMO equalizer and the ICSI-aware SIC error regularization and present its application in a simulation for a Turbo-MIMO iterative receiver with feedback of a-posteriori information from the channel decoder to the symbol detector