Iterative Detection for MIMO Systems Associated with Polar Codes

The tools of information theory (error correcting codes) are exploited to improve the detection of MIMO channels. In particular, we use the polar codes, which currently represent the first and only code family to reach Shannon's two limits on channel and source coding. Our proposal consists in using a medium-length polar encoding, a detection minimizing the mean squared error by successive interference cancellation (MMSE-SIC) and a Successive Cancellation (SC) decoder. To improve the reliability of the information over the iterations, we added an iterative (turbo-equalization) process between the SC decoder and the MMSE-SIC detector. We have shown that the association of the SC decoder with the MMSE-SIC detector can reduce the Bit Error Rate (BER) instead of Believe Propagation (BP) decoding. This has increased the diversity and the ability to correct.