
Entropy, Год журнала: 2024, Номер 26(12), С. 1056 - 1056
Опубликована: Дек. 5, 2024
This paper investigates the design of active and passive beamforming in a reconfigurable intelligent surface (RIS)-aided multi-user multiple-input single-output (MU-MISO) system with objective maximizing sum rate. We propose deep evolution policy (DEP)-based algorithm to derive optimal strategy by generating multiple agents, each utilizing distinct neural networks (DNNs). Additionally, random subspace selection (RSS) is incorporated effectively balance exploitation exploration. The proposed DEP-based operates without need for alternating iterations, gradient descent, or backpropagation, enabling simultaneous optimization both beamforming. Simulation results indicate that can bring significant performance enhancements.
Язык: Английский