Performance analysis of linear precoding based on field trials results of MIMO-OFDM system
Sampath, H.
Erceg, V.
Paulraj, A.
Qualcomm Inc., San Diego, CA, USA;
This paper appears in: Wireless Communications, IEEE Transactions on
Publication Date: March 2005
Volume: 4,
Issue: 2
On page(s): 404- 409
ISSN: 1536-1276
INSPEC Accession Number: 8354275
Digital Object Identifier: 10.1109/TWC.2005.844151
Current Version Published: 2005-04-04
Abstract
We use field trial results obtained from a multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) wireless system with two transmitter and three receiver antennas (2×3), to first validate the properties of the transmit correlation matrix in a macro-cellular environment. We find that approximately 20% of the locations have well-defined transmit correlation matrices. Furthermore, the eigenvectors of the transmit correlation matrix vary slowly over distance with 60% of the locations having eigenvector variation of less than 1 dB over a distance of 20 m. Next, we quantify the performance of the optimal statistical linear precoding (OSLP) , and statistical one-dimensional (1-D) eigenbeamforming (SEB) based on transmit correlation matrices, and the 1-D eigenbeamforming (EB)-based on perfect channel knowledge at the transmitter. We find that the OSLP and SEB schemes obtain array gain over the Alamouti scheme at lower signal-to-noise ratio (SNR) with a median gain of 2.0 (1.5) dB at the 1.0-(3.5) km cell-radii. However, the SEB scheme (unlike the OSLP scheme) looses diversity order at higher SNR that leads to a performance loss. The EB scheme provides the best performance over the Alamouti scheme, at the expense of increased feedback requirements.
Index
Terms
Available to subscribers and IEEE members.
References
Available to subscribers and IEEE members.
Citing Documents
Available to subscribers and IEEE members.
Cart
