Sphere-decoding aided SIC for MIMO-GFDM: Coded performance analysis

Maximilian Matthé, Dan Zhang, Gerhard P. Fettweis

Abstract: Multiple-input multiple-output (MIMO) detection algorithms are of great importance for high-performance mobile communications and future waveforms need to be MIMO capable. Generalized Frequency Division Multiplexing (GFDM), as a promising non-orthogonal waveform, includes 3-dimensional interference which was shown to be beneficial in terms of exploiting time, frequency and space diversity for improved performance while iterative receivers have been presented to attain this gain. This paper considers non-iterative receiver design. Without relying on the feedback information from the decoder, we improve a previously proposed sphere decoding (SD) based MIMO-GFDM detection algorithm and analyze its performance in the coded case. The results show that the proposed algorithm cannot compete with SD based detection for OFDM because of intersymbol interference (ISI) and intercarrier interference (ICI) in addition to interantenna interference (IAI). These observations indicate that considering the GFDM MIMO detector as a joint operation between signal demapper and channel decoder is more promising for exploiting the foreseen capacity gains than trying to optimize the functional units separately.