Deep-Learning-Assisted IoT-Based RIS for Cooperative Communications

Reconfigurable intelligent surfaces (RISs) are software-controlled passive devices that can be used as relay (R) systems to reflect incoming signals from a source (S) destination (D) in cooperative manner with optimum signal strength improve the performance of wireless communication networks. The configurability and flexibility an RIS deployed Internet-of-Things (IoT)-based network enable designers devise stand-alone or configurations have considerable advantages over conventional In this paper, two new deep neural (DNN)-assisted models, namely, DNNR-CRIS DNNR,D-CRIS, proposed for communications. model, potential deployment IoT-based element next-generation is investigated using learning (DL) techniques phase optimization. addition, reduce maximum likelihood (ML) complexity at D, DNN-based symbol detection method presented DNNR,D-CRIS model combined DNN-assisted For different number relays receiver configurations, bit error rate (BER) results models traditional (CRIS) scheme (without DNN) multi-relay scenario path loss effects. It revealed show promising terms BER, even high-noise environments low system complexity.