Reduced complexity MIMO detectors for LDPC coded systems

We consider the performance and complexity analysis of low-density parity-check (LDPC) coded multiple-input multiple-output (MIMO) systems with reduced complexity detectors in fast Rayleigh fading channels. We use a turbo iterative detection-and-decoding receiver which combines a soft MIMO demodulator and a message-passing LDPC decoder We examine three low complexity detectors: minimum mean-square error (MMSE) with soft interference cancellation (SIC), MMSE suppression, and MMSE with hard decision interference cancellation (MMSE-HIC). We compare the performance and complexity of these detectors with the standard soft maximum a posteriori (MAP) detector and show that they provide a lower computational complexity with only a small penalty in performance (∼0.5 dB). We also provide comparisons with the Shannon capacity limits for ergodic multiple-antenna channels and show that an LDPC coded system using a length 15,000 code and QPSK modulation with reduced-complexity detection operates 2 and 3 dB from the ergodic capacity of the Rayleigh MIMO channel.