Optimized RIS-Assisted Joint Multi-Domain Index Modulation for OTFS via Deep Learning

Reconfigurable intelligent surfaces (RIS) have emerged as a key technology in wireless communications due to their ability to actively manipulate the wireless channel environment, thereby enhancing communication quality. This paper proposes a novel RIS-assisted joint multi-domain index modulation (IM) orthogonal time frequency space (OTFS) system termed RIS-JMDIM-OTFS. This system leverages transmit antenna, receive antenna, and delay-Doppler (DD) indices within the OTFS framework to achieve higher transmission efficiency. Additionally, it exploits the electromagnetic properties of the RIS to optimize the wireless channel, improving bit error rate (BER) performance. To reduce receiver hardware costs and energy overhead, a deep neural network-based bit sequence detection (DNN-BSD) method is introduced, which effectively reduces the computational complexity associated with high modulation orders during detection. Simulation results demonstrate that the RIS-JMDIM-OTFS system outperforms state-of-the-art IM OTFS system in terms in both transmission efficiency and BER performance. Furthermore, DNN-BSD achieves near-identical detection performance to maximum likelihood (ML) detection while significantly reducing computational complexity.