Object

Group publication title:

2025, nr 4, JTIT-artykuły

Description:

kwartalnik

Abstrakt:

Millimeter-wave (mmWave) communication is a key enabler of 5G and future wireless systems, providing vast bandwidth for high-speed data transfers. However, high user mobility leads to significant Doppler shifts, which can severely degrade the performance of beamforming - an essential technology for mmWave systems. The traditional hybrid beamforming (HBF) technique faces challenges in adapting to rapid channel variations caused by Doppler effects. Therefore, this paper introduces a deep learning framework to mitigate Doppler-induced channel distortions in hybrid beamforming. We propose a long-short-term memory (LSTM)-based neural network that predicts Doppler shifts and dynamically adjusts the hybrid beamforming vectors to compensate for these variations. This approach proactively addresses channel distortion, enhancing both spectral and energy efficiency. The simulation results and the performance comparison of proposed model against conventional beamforming and state-of-the-art techniques demonstrate the superiority of deep learning-based solution in maintaining robust communication links under high-mobility conditions, showcasing its potential to improve performance in next-generation wireless networks.

Number:

4

Publisher:

National Institute of Telecommunications

Resource Identifier:

oai:bc.itl.waw.pl:2421

DOI:

10.26636/jtit.2025.4.2349

eISSN:

on-line: ISSN 1899-8852

Source:

Journal of Telecommunications and Information Technology

Language:

ang

Rights Management:

Biblioteka Naukowa Instytutu Łączności

License:

CC BY 4.0