Noise Immunity Analysis of Massive MIMO Demodulation Algorithms Using Non-Gaussian Approximation

Massive MIMO technology is one of the fundamental technologies in the development of modern wireless communications (5G, B5G, Wi-Fi 7), which is confirmed by a large number of scientific papers in this area, but despite all the achievements, in terms of strict and extremely high requirements for speed and quality of communication, there is undoubtedly an urgent need for highly efficient demodulation algorithms with low computational complexity. The paper presents an extensive analysis of the noise immunity of various demodulation algorithms for different numbers of antennas, the QAM modulation order, the noise-resistant turbo coding rate, and the non-Gaussian approximation order. The results of this study make it possible to optimize the operation of demodulation algorithms using non-Gaussian approximation of the a priori distribution, taking into account the parameters of the Massive MIMO system, the type of modulation, and the encoding rate. The adequacy of the results is confirmed by joint modeling and comparison with well-known algorithms such as MMSE and K-best, under different conditions.