Title
An efficient and scalable hardware architecture for singular value decomposition towards massive MIMO communications
Abstract
Massive multiple input multiple output (MIMO) technology plays an important role in next generation wireless communication systems. Modified Brent-Luk-Van Loan array and other parallel hardware implementations were developed for channel matrix factorization. For a large matrix size as of massive MIMO, however, previous implementations would require a large amount of hardware resource. This paper presents a hardware-efficient architecture that performs singular value decomposition of a complex matrix in arbitrary size. Eveluated on an FPGA platform, the proposed architecture results faster or competitive processing time for matrix factorization but consumes significantly less hardware resource.
Year
DOI
Venue
2017
10.1109/MWSCAS.2017.8053011
Midwest Symposium on Circuits and Systems Conference Proceedings
Keywords
Field
DocType
massive multiple input multiple output (MIMO),singular value decomposition,hardware-efficient architecture,scalable order
Singular value decomposition,Matrix (mathematics),Computer science,Matrix decomposition,Communication channel,Field-programmable gate array,MIMO,Electronic engineering,Computer engineering,Scalability,Distributed computing,Hardware architecture
Conference
ISSN
Citations 
PageRank 
1548-3746
0
0.34
References 
Authors
8
4
Name
Order
Citations
PageRank
Mingda Zhou132.46
Youjian Liu260549.82
Tian Xia3395.50
Xin-ming Huang435646.91