Min-Sum Soft Symbol Estimation for Iterative MMSE Detection

Soft symbol estimation (SSE) is the first process in soft interference cancellation minimum mean squared error (SIC-MMSE) detection for multiple-input multiple-output (MIMO) systems. SSE requires the sum of exhaustive multiplications of probability that occupies a non-negligible amount of the entire SIC-MMSE complexity. This paper proposes two approaches to reduce the complexity of SSE. The first is to find the approximation of SSE by investigating the estimation in the log-domain, which leads to a simple min-sum operation. The second is to approximate the hyperbolic tangent with low-order piecewise polynomial functions. These two approaches are then integrated. The simulation results show that the proposed methods do not degrade the performance in terms of bit error rate (BER), and that they can greatly reduce the number of multiplications, from O(K2K) to O(K).