Using Deep Learning To Investigate Massive MIMO Hybrid Beamforming Design
| dc.contributor.author | khodja et Guelmine, safa et Amel | |
| dc.date.accessioned | 2021-07-18T12:42:02Z | |
| dc.date.available | 2021-07-18T12:42:02Z | |
| dc.date.issued | 2021 | |
| dc.description.abstract | MIMO antenna has offered a promising potential to speed up the 5G evolution. However, the number of antenna arrays and resource allocations lead to decreasing the spectral efficiency or the capacity of 5G networks. To deal with these issues, the hybrid precoder and combiner design has been proposed to alleviate the computation complexity related to the number of fully digital beamformers selected. This dissertation examines the design of the hybrid precoder and combiner (HPC) in mm-Waves by integrating a deep learning algorithm. In particular, we have investigated the potentiality of the CNN algorithm on the estimation of the analog precoders and combiners parameters respectively. Results obtained from simulations demonstrate the improved performance of the CNN-MIMO algorithm in contrast to the existing algorithms and can achieve promising results. | en_US |
| dc.identifier.uri | http://dspace.univ-msila.dz:8080//xmlui/handle/123456789/25134 | |
| dc.language.iso | en | en_US |
| dc.publisher | FACULTY :Mathematics And Computer Science DOMAIN : Mathematics And Computer Science DEPARTMENT : Computer Science | en_US |
| dc.subject | Hybrid precoding/combining, mm-Wave, SNR, Spectral efficiency, Deep Learning. | en_US |
| dc.title | Using Deep Learning To Investigate Massive MIMO Hybrid Beamforming Design | en_US |
| dc.type | Thesis | en_US |