Multiuser MIMO Downlink Beamforming with Finite Rate Channel Feedback
Date Issued
2009
Date
2009
Author(s)
Chen, Chun-Yao
Abstract
Multi-user Multi-Input Multi-Output (MIMO) downlink beamforming problem with individual signal-to-interference-plus-noise ratio (SINR) constraints is considered. We consider a general algorithm for this problem, which has great performance, but needs perfect channel state information at transmitter (CSIT). For practical considerations, we consider a system with finite rate channel feedback. In this case each receiver measures perfect and instantaneous channel state information (CSI)nd quantizes it with a finite number of bits. The transmitter receives the quantized CSI to design the beamforming filters and power allocation.e first compress the CSI for this algorithm. A new channel matrix is generated by singular value decomposition. By this operation, the new CSI can be better quantized, and it is proved that feeding back the new CSI is equivalent to feeding back the original one. Secondly, based on this observation, a quantization scheme for the new CSI is proposed. We decompose the rotated channel matrix into twoarts to quantize separately. As a result, we can adjust the quantization bit allocation between the two parts, according to the varying communication environment. On one hand, one can quantize the CSI more efficiently with a fixed number of bits. On the other hand, we show that linearly scaling the number of feedback bits on only one of the two parts is su?cient to maintain a bounded system performance loss.
Subjects
multi-user Multi-Input Multi-Output (MIMO)
downlink channel
beamforming
Inter-User Interference (IUI)
finite rate channel feedback
codebook based quantization
random quantization codebook
ceiling effect
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-98-R96942049-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):591364f16599cc6368091fd0bd606066