Matrix Manifold Precoder Design for Massive MIMO Downlink

We investigate the weighted sum-rate (WSR) max-imization linear precoder design under total power constraint (TPC) for massive MIMO downlink with matrix manifold optimization. Particularly, we prove that the precoders under TPC are on a Riemannian submanifold, and transform the constrained problem i...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE Wireless Communications and Networking Conference : [proceedings] : WCNC s. 1 - 6
Hlavní autoři: Sun, Rui, Wang, Chen, Lu, An-An, Fu, Xiao, Liu, Xiaofeng, Zhang, Yuxuan, Gao, Xiqi, Xia, Xiang-Gen
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 21.04.2024
Témata:
Computational modeling
Design methodology
Downlink
linear precoding
Manifolds
Massive MIMO
matrix manifold optimization
Simulation
total power constraint
Transforms
weighted sum rate
ISSN:1558-2612
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We investigate the weighted sum-rate (WSR) max-imization linear precoder design under total power constraint (TPC) for massive MIMO downlink with matrix manifold optimization. Particularly, we prove that the precoders under TPC are on a Riemannian submanifold, and transform the constrained problem in Euclidean space to the unconstrained one on manifold. In accordance with this, Riemannian design methods using Riemannian steepest descent and Riemannian conjugate gradient are provided to design the WSR-maximization precoders under TPC. Riemannian methods are free of the inverse of large dimensional matrix, posing significant computational savings and potentially allowing to avoid ill numerical behavior in algorithms. Complexity analysis and performance simulations demonstrate the advantages of the proposed precoder design.
ISSN:1558-2612
DOI:10.1109/WCNC57260.2024.10570510