On the Design of Near-Optimal Generalized Block-Based Spatial Modulation With Low Detection Complexity

Spatial Modulation (SM) and Generalized Spatial Modulation (GSM) have attracted significant attention in the development of spectrally and energy-efficient transmission schemes for multiple-input multiple-output (MIMO) systems. However, independently designing the constellation cardinality and TACs leads to limited performance gains and an exponential increase in complexity, particularly under maximum-likelihood (ML) detection. To address these limitations, the generalized block-based spatial modulation (GBSM) scheme was proposed, enabling greater flexibility by jointly designing GSM signals across a block of time indices. Building on this idea, this paper first proposes a near-optimal codebook search method based on three-dimensional (3-D) mapping, applicable to both fast and slow Rayleigh fading channels. Secondly, a codebook-assisted tree-search detector (CATSD) is introduced, offering a 98% reduction in complexity compared to ML detection while maintaining near-ML error performance. Finally, an alternative codebook search method is presented, accompanied by a complexity analysis that reveals a favorable trade-off between performance and computational cost.