Date of Award
Spring 2026
Language
English
Embargo Period
4-7-2026
Document Type
Master's Thesis
Degree Name
Master of Science (MS)
College/School/Department
Department of Electrical and Computer Engineering
Program
Electrical and Computer Engineering
First Advisor
Aveek Dutta
Committee Members
Aveek Dutta, Dola Saha, Bariscan Yonel
Keywords
LTV Channels, Neural Networks, MEM, Decomposition, ALM
Subject Categories
Systems and Communications
Abstract
Eigenfunctions are commonly employed to characterize kernels in various data-driven analyses. In machine learning, eigenfunction decomposition typically relies on Mercer's theorem, which assumes kernel symmetry. However, this condition is often unmet in communication systems, where channel kernels are asymmetric due to differences in downlink and uplink propagation environments. The High-Order Generalized Mercer's Theorem (HOGMT) provides a systematic approach for decomposing multidimensional asymmetric kernels into eigenfunctions. To address the complexity of eigen-decomposition, this work introduces a baseline neural network (NN) framework HNET. The HNET framework is further enhanced by incorporating the augmented Lagrangian method (ALM) to explicitly enforce orthogonality constraints. This ALM-based NN framework AHNET reduces the training time by eliminating the need for extensive parameter tuning. The ongoing work proposes a Cloud Radio Access Network (CRAN), a hardware platform to facilitate the evaluation of a range of research directions, including machine learning based wireless communication.
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Amarasekara, Iresha, "Neural Network Transceiver for LTV MIMO Channels" (2026). Electronic Theses & Dissertations (2024 - present). 397.
https://scholarsarchive.library.albany.edu/etd/397