Hierarchical Multi-Functional Layered Spatial Modulation

In pursuit of optimal index modulation -aided multiple-input multiple-output (MIMO) systems, where information is implicitly conveyed by relying on the on/off mechanism of the system's components in addition to the classical amplitude, phase, or frequency components, we present in a tutorial style our novel multi-functional (MF) architecture of layered multi-set (LMS) modulation. This generalized framework subsumes various MIMO techniques exhibiting different multiplexing and diversity functionalities. Our LMS design relies on three constituents, namely the space-time (ST) unit, the layered unit and the spatial switching unit. More specifically, the ST unit relies on the generalized space-time shift keying (GSTSK) scheme, where P - rather than one - out of Q ST dispersion matrices are selected for dispersing an equivalent number of phase-shift keying/quadrature amplitude modulation symbols across the antennas and timeslots. In the layered unit, multiple GSTSK codewords are stacked within the layers of codewords spread over time and space. The spatial switching unit activates N
_c
^t
out of N
_t
transmit antennas. Owing to its hierarchical MF architecture, our LMS system strikes a flexible design trade-off between the achievable throughput as well as the attainable diversity gain and it can potentially subsume various conventional MIMO schemes, such as Bell Lab's Layered Space-Time, space-time block codes, layered steered space-time codes, spatial modulation (SM), space-shift keying, linear dispersion codes, generalized SM, STSK, GSTSK, quadrature SM and multi-set STSK . Additionally, we derive the LMS system's discrete-input continuous-output memoryless channel capacity, which encompasses the capacity limit of all the LMS subsidiaries. We also propose a two-stage serially concatenated soft-decision (SD) based LMS detector by relying on an inner and an outer decoder that iteratively exchange their extrinsic information in order to achieve a near-capacity performance. Last but not least, we utilize the extrinsic information transfer charts for analyzing the convergence behavior of our SD-aided coded LMS scheme.