Downlink transmission rate-control strategies for closed-loop multiple-input multiple-output systems

Dept. of Electr. & Comput. Eng., Univ. of Texas at Austin, Austin, TX
IET Communications (Impact Factor: 0.74). 05/2009; DOI: 10.1049/iet-com.2008.0456
Source: IEEE Xplore

ABSTRACT A novel downlink transmission rate-control and feedback reduction strategy for closed-loop multiple-input multiple-output (MIMO) multiple-input multiple-output wireless systems is presented. Unlike conventional systems that use signal to interference plus noise ratio at the receiver as an indicator of channel quality, we propose using instantaneous MIMO capacity as an indicator for the downlink transmission rate-control. A set of instantaneous capacity thresholds is first chosen such that the expected weighted capacity loss because of thresholding effects are minimised. While computing the thresholds, we also consider the quality of service and weight function to meet different traffics and user needs. Then a set of codebooks can be constructed minimising the overall capacity loss with given quality of service constraint. Simulation results show that, with only four data rate-control bits, our algorithm gives only 12% capacity loss in 4 times 4 MIMO systems and almost twice better than the current IS-856 standard in single-input single-output systems. In case of 5-bit feedback scenario, the proposed algorithm outperforms conventional systems by minimising instantaneous capacity loss.

1 Follower
  • [Show abstract] [Hide abstract]
    ABSTRACT: An opportunistic feedback scheme is proposed for reducing the uplink control information feedback for the coordinated scheduling/beamforming (CS/CB) mode in the coordinated multiple point (CoMP) systems. CoMP is a promising technology for LTE-Advanced to improve the coverage and the spectral efficiency. However, the total feedback load is substantially heavy and becomes a serious problem for the practical systems. The proposed opportunistic feedback scheme can reduce the feedback load with little user throughput deterioration. In the proposed feedback scheme, in each cooperating cells set, the cells are divided into one primary cell and several assistant cells. Users in the primary cell are fairly determined to feed back according to the normalized signal to interference plus noise ratio (SINR) threshold. In the assistant cells, users who cause less interference to the primary cell are supposed to feed back. The simulation results show that the proposed scheme reduces the feedback load magnificently, while keeping the performance loss minimal.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Water-filled eigenchannels offer the highest multi-input multi-output (MIMO) information-theoretic capacity, but digital techniques such as quadrature amplitude modulation and finite block lengths will degrade the capacity from the Shannon limit to the capacity of a digital link. Furthermore, eigen-MIMO requires channel overheads, such as estimating the channel state information (CSI) and feeding it back to the transmitter, which further compromise the capacity. In this paper, the joint influence of channel estimation and imperfect feedback on the information-theoretic capacity and the practicable capacity is analyzed. The channel is modeled as static over a MIMO channel block. In each block, the forward channel is used for CSI estimation and for the payload data transmission. In the back direction, the channel is used to feed back a quantized form of the CSI to the transmitter with a throughput constraint. These three channel usages are combined into an effective simplex channel simplifying the capacity analysis. The capacities are formulated as functions of the link parameters, enabling optimization of the number of training symbols, the feedback duration, and the power allocation for training and data transfer, with the criterion of maximum capacity. The results presented are subject to the usual approximations used in communications theory. Copyright © 2012 John Wiley & Sons, Ltd.
    Wireless Communications and Mobile Computing 11/2014; 14(16). DOI:10.1002/wcm.2297 · 1.29 Impact Factor