Francisco Rubén Castillo SoriaRoilhi Frajo Ibarra‐HernándezCarlos Adrián Gutiérrez Diaz de LeónAbel García BarrientosSharon Macias‐VelasquezDel-Puerto-Flores, J. AlbertoJ. AlbertoDel-Puerto-Flores2025-07-022025-07-022025-05-1697813942501109781394250141https://scripta.up.edu.mx/handle/20.500.12552/1221710.1002/9781394250141.ch5Integrating reconfigurable intelligent surfaces (RISs) into wireless communication systems represents a crucial challenge. This task can be even more challenging when dealing with multiple users connected to a system sharing spectrum, time, or power resources. This chapter explores the challenges of integrating the RIS into multiuser downlink transmission systems. The efficiency and applicability of RIS-assisted multiple input-single output (MISO) and multiple input-multiple output (MIMO) schemes are analyzed in terms of bit error rate (BER) performance and complexity of the algorithms and techniques utilized. Likewise, the effects of the different wireless propagation channel models are analyzed. Approaches of blind RISs and optimization algorithms are also reviewed. Our simulation results showed up to 37 dB gain in BER curves when using N-RIS surfaces when the system has SE = 12 bpcu/user, 32 Tx antennas, and 8 users with 4 Rx antennas. These outcomes validate an increase in performance by adopting N-RIS surfaces and optimization techniques for adequate phase searching. This chapter also reviews the reported frameworks that apply machine learning algorithms to improve the overall system performance. Topics such as estimation of channel state information, beamforming applications, federated learning, and demodulation applications are reviewed. Finally, we shed light on trends and open research areas in this emerging topic. © 2024 The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.en-USAcceo Restringido.multiple RISMU-MIMO systemsmachine learningtext::book::book part