Skripsi

OPTIMASI BEAMFORMING SINYAL ANTENA MIMO PADA JARINGAN 5G MENGGUNAKAN ALGORITMA LONG SHORT-TERM MEMORY (LSTM)

---

This study discusses beamforming optimization in indoor 5G communication systems based on MU-MIMO configuration by comparing two main approaches, namely the conventional adaptive Least Mean Square (LMS) algorithm and the Long Short-Term Memory (LSTM) deep learning-based weight prediction method. The tested channel environment is an indoor scenario with ten random users, resulting in complex multipath channels, inter-user interference, and Angle of Arrival (AoA) variations. The LSTM model was trained using phase-normalized MVDR weights and 36 channel features covering spatial information, CSI, and interference. Ten position scenarios were tested with SINR, throughput, and latency parameter analysis for each user. The test results show that LSTM consistently provides performance improvements for most user equipment (UE), especially on channels with high interference, characterized by more precise main lobes, lower sidelobes, and higher SINR and throughput. LMS only excels on UE with isolated AoA or relatively clean channels. Overall, LSTM proves to be more stable and effective in handling complex multiuser channels.

---

Availability

Detail Information

Series Title

-

Call Number

T1901602025

Publisher

Indralaya : Prodi Teknik Elektro, Fakultas Teknik Universitas Sriwijaya., 2025

Collation

xx, 253 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

621.384 507

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Elektro
Jaringan Seluler 5G

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment