| Full text | |
| Author(s): |
de Souza, Pedro H. C.
;
Mendes, Luciano L.
Total Authors: 2
|
| Document type: | Journal article |
| Source: | 2023 IEEE CONFERENCE ON STANDARDS FOR COMMUNICATIONS AND NETWORKING, CSCN; v. N/A, p. 4-pg., 2023-01-01. |
| Abstract | |
The search for enabling technologies that will be part of the next generation of mobile communications is at a crucial point. The RISs, for instance, are passive metasurfaces that can be configured with a aim at improving the channel propagation conditions. This is done by optimizing the configuration of its reflectors or elements, such that the channel achievable rate is maximized, for example. We focus on the RIS performance under the practical assumption that a subset of its elements can be at fault. Moreover, we also propose a neural network (NN) architecture for configuring the RIS and compare it to the seminal strongest tap maximization (STM) method. Numerical results show a considerable performance loss when increasing the ratio of faulty elements in the RIS. It is also demonstrated that the NN requires fewer reconfigurations of the RIS. (AU) | |
| FAPESP's process: | 20/05127-2 - SAMURAI: smart 5G core and multiran integration |
| Grantee: | Aldebaro Barreto da Rocha Klautau Junior |
| Support Opportunities: | Research Projects - Thematic Grants |