MRC Detection for LDPC-OFDM, Massive MIMO, NR-5G-based Systems Utilizing Accurate PDF of Effective Noise and Co-Channel Interference

This paper proposes a technique to improve bit error rate (BER) performance of soft-output maximal-ratio combining (MRC) detection for Low-density parity-checked (LDPC)-coded, orthogonal frequency-division multiplexing (OFDM) based massive multiple-input, multiple-output (MMIMO) systems. The detector uses a derived probability distribution function (PDF) instead of the conventional Gaussian distribution to produce soft information in form of the log-likelihood ratio (LLR) for iterative decoders. In order to minimize the complexity of the detector, we utilize a polynomial equation using Newton’s method to produce approximate LLRs. The BER performance over frequency-selective, multi-path, Rayleigh fading channels was investigated through simulation, and the results confirm that the proposed detector provided a better BER performance than that of the classical, low-complexity, soft-output detections, especially at the high E
b
/N
0
region. Therefore, the proposed detection is a powerful technique for improving the reliability of LDPC-OFDM-MMIMO systems.