Abstract
Linear dispersion codes (LDCs) are an attractive tool for attaining spatial diversity in OFDM systems with multiple transmit antennas. To reach the potential gains, in each situation, the selection of a proper set of codes depends, in general, on the existence of a reliable feedback channel. We propose LDCs for multiple-input multiple-output (MIMO) channels with a fixed amount of feedback. The proposed scheme selects the LDC from a set of LDCs that minimizes the error probability based on the instantaneous channel conditions. The determination of the best set of LDCs is described as a constrained optimization problem. We present several good sets of LDCs, obtained from an optimization algorithm. Through simulation results we show that the proposed schemes outperform previously reported comparable results, yet at the price of increased complexity. We also show the performance gains that can be reached in practice with LDC, by critically evaluating the feedback channel impairments, such as errors and signalling delays, and also considering the effect of signalling overhead on the achievable throughput.
Highlights
M ULTIPLE-INPUT MULTIPLE-OUTPUT (MIMO) wireless communication systems can exploit the spatial dimension to achieve higher transmission rates and/or to improve the error performance over fading channels
One way of obtaining the benefits of the spatial dimension for wireless channels with multiple transmit antennas is through spacetime coding [1], which requires no knowledge of the channel state information (CSI), i.e., the fading coefficients associated with the channels from the transmit to the receive antennas, at the transmitter
For MISO or MIMO wireless channels, the performance can be considerably improved if a feedback channel exists so that the CSI is known at the transmitter
Summary
M ULTIPLE-INPUT MULTIPLE-OUTPUT (MIMO) wireless communication systems can exploit the spatial dimension to achieve higher transmission rates and/or to improve the error performance over fading channels. For wireless channels with MT ≥ 2 transmit antennas, MR receive antennas, and limited feedback, we propose a code selection (CS) transmission scheme that uses the best code from a set of 2bf codes, where bf is the number of feedback bits. Since for some parameters the optimization algorithm outputs the uniform Q-EGT solution, the results in this paper support the claim made in [7, Section VII] that the uniform quantization is very likely to be the optimum scalar quantized when the total number of feedback bits is a multiple of (MT − 1) Those LDCs were not studied in more realistic conditions where limitations of the feedback channel are present.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
