Belief propagation detection based on max-sum algorithm for massive MIMO systems

In this paper, belief propagation (BP) detection based on max-sum (MS) algorithm for massive multiple-input multiple-output (MIMO) systems is therefore proposed to reduce computational complexity of general belief propagation. Owing to employing the approximation strategy, complexity reduction of MS is at the expense of detection performance loss. Based on MS algorithm, two effective approaches are proposed to compensate the performance loss resulting from MS detection. By introducing a normalized factor or an offset factor, performance and complexity can strike a balance to some extent for large-scale MIMO systems. Simulation results show that, for asymmetric antenna configuration, MS detection suffers negligible performance loss while keeping lower hardware complexity compared to original BP detection. Furthermore, in comparison with MS algorithm, the two modified algorithms can achieve noticeable performance improvement with minor overhead hardware complexity for symmetrical antenna configurations, which is applicable to independent identically distributed (i.i.d.) and correlated channels.