Prioritized Adaptive Modulation for MIMO-OFDM Using Pre-Ordered SIC

In MIMO transmission, channel state information (CSI) is crucial for achieving channel adaptation. However, the inaccuracy of CSI may induce severe interferences. Hereto, limitations of linear equalizers to combat severe interference and noise enhancements necessitate the need for investigating non-linear schemes. Thus, we propose a modified successive interference cancellation (SIC) technique based on the well-known V-BLAST non-linear spatial equalizer. First, we implement a linear pre-processing filter in order to pre-sort the eigenchannels at the transmitter. This simplifies the complexity of non-linear equalization significantly by reducing the effort needed for sorting at the receiver. To protect the strong eigenbeams against errors and minimize the SIC error propagation, an unequal-error protection (UEP) bit-loading algorithm is used. A comparison to an MMSE linear equalization shows that our design operates at a lower symbol-error ratio (SER) with almost identical complexity.