Spherical Linear Interpolation for Transmit Beamforming in MIMO-OFDM Systems with Limited Feedback
Abstract
Transmit beamforming with receive combining is a simple approach to exploiting the significant diversity provided by multiple-input multiple-output (MIMO) systems and this technique can be easily extended to frequency selective MIMO channels by employing orthogonal frequency division multiplexing (OFDM). Optimal beamforming requires channel state information in the form of the beamforming vectors corresponding to all the OFDM subcarriers. When the uplink and downlink channels are not reciprocal this information must be conveyed back to the transmitter. To reduce the amount of feedback information a new approach to transmit beamforming is proposed that combines limited feedback and beamformer interpolation. Because the length of the OFDM cyclic prefix is designed to be much less than the number of subcarriers to increase spectral efficiency the neighboring subchannels of a MIMO-OFDM system are substantially correlated. Thus the beamforming vectors determined by the subchannels are also significantly correlated. To reduce the feedback information using the correlation between beamforming vectors the receiver of the proposed scheme sends back only a fraction of information about the optimal beamforming vectors to the transmitter. Then the transmitter evaluates the beamforming vectors for all subcarriers through interpolation of the conveyed beamforming vectors. Since a beamforming vector is phase invariant and has unit norm a new spherical linear interpolator is proposed that exploits additional parameters for phase rotation. These parameters are determined at the receiver in the sense of maximizing the minimum channel gain or capacity and they are sent back to the transmitter along with the beamforming vectors through the feedback channel.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 20, 2004
- Accession Number
- ADA432697
Entities
People
- Jihoon Choi
- Robert W. Heath Jr.
Organizations
- University of Texas at Austin