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 time-slots. 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.
Read full abstract