Towards complexity-reduced Soft-Input Soft-Output Sphere Detection

Iterative detection?decoding is a powerful approach for wireless multiple-input multiple output (MIMO) transmission to achieve very high detection performance, but comes at the costs of high processing complexity at the receiver side. In order to enable efficient hardware realizations, it is necessary to suitably reduce the inherent increase in complexity when a-priori information (i.e. soft-input) is taken into account in the detection process. In this paper, a probabilistic analysis is described and carried out to gain profound knowledge of softinput soft-output (SISO) detection, which can be further exploited to develop new strategies for complexity reduction. Based on this analysis, two novel approaches with very low computational complexity are presented, their application results in efficient SISO detection algorithms, well suited for implementation.