The emerging technology of reconfigurable intelligent surface-based access points (RIS-APs) shows excellent potential as it effectively replaces conventional RF front ends with passive RISs. This paper introduces RIS-based dual-mode receive spatial modulation (RDM-RSM), a low-complexity scheme designed to improve the spectral efficiency and error performance of RIS-APs. Both versions of the proposed RDM-RSM scheme configure the RIS to simultaneously enhance the RF signals at two different receive antennas while synthesizing distinct modulation symbols from two separate constellations at each antenna, thereby enabling efficient transmission of two parallel, conventionally-modulated data-streams. Additionally, by selecting two antennas and assigning each a distinct constellation mode, more bits can be encoded per index-modulated symbol. The first version, RDM-RSM-Independent (RDM-RSMI), operates by maximizing the energy of the synthesized signalsindependentlyfor each antenna. In contrast, RDM-RSM-Joint (RDM-RSMJ) optimizes the reflection phase of each RIS element to maximize an objective function that considers the signals at both antennasjointly, resulting in stronger received signals compared to RDM-RSMI, where only half of the elements can serve each antenna. A low-complexity detector is proposed, with its theoretical performance rigorously analyzed by modeling the conditional reflection phases of the RIS elements as von-Mises distributed, providing a tractable and insightful framework.

Besides steering the incoming signals toward a receiver, a reconfigurable intelligent surface (RIS) can simultaneously serve as a backscatter node to communicate additional data by modulating the same signal. Quadrature reflection modulation (QRM) is a promising contender among index modulation (IM)-based techniques for implementing RIS-based backscatter communications (RIS-BackCom). In this paper, we demonstrate that the error performance of RIS-QRM heavily relies on the employed indexing scheme, which configures each RIS element into one of the two orthogonally-reflecting modes of QRM to encode the transmitted data onto the reflection pattern of RIS elements. In this context, we present an efficient guideline for designing indexing schemes that ensure the resulting channel-dependent QRM symbols are evenly spaced with maximum separation. Following this guideline, we propose two specific indexing techniques: the incremental and uniform indexing schemes, both of which substantially enhance the error performance of RIS-QRM without increasing detection complexity. We derive a closed-form expression for the bit error probability (BEP) of RIS-QRM with the incremental indexing scheme. Furthermore, we propose a low-complexity detector for RIS-BackCom systems employing RIS-QRM. Extensive computer simulations illustrate the effectiveness of our proposed schemes in improving the error performance of RIS-QRM-based BackCom systems.