Low-Complexity FNN-Based Transmit Antenna Selec tion for Enhanced Performance in VASM Systems over Rician Fading Channels
Từ khóa
DOI:
https://doi.org/10.32913/mic-ict-research.v2025.n1.1346Tóm tắt
Variable Active Antenna Spatial Modulation (VASM) is a spatial modulation variant designed to improve spectral efficiency and offer greater flexibility in system configuration. In this paper, we propose a feedforward neural network (FNN) framework to address the transmit antenna selection (TAS) problem, aiming to enhance performance in VASM systems operating over Rician fading channels. Computational results demonstrate that our novel FNN-TAS algorithm provides a significant reduction in computational costs compared to the standard Euclidean distance-based method. Additionally, simulation results indicate that the bit error rate (BER) of the VASM system increases with the Rician factor. However, the proposed FNN-TAS method effectively improves BER performance for VASM systems,
particularly at high Rician factors, and outperforms conven tional TAS methods based on channel gain.