By Topic

Worst-Case Robust MIMO Transmission With Imperfect Channel Knowledge

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Jiaheng Wang ; Dept. of Electron. & Comput. Eng., Hong Kong Univ. of Sci. & Technol., Hong Kong, China ; Daniel P. Palomar

In this paper, we consider robust transmit strategies, against the imperfectness of the channel state information at the transmitter (CSIT), for multi-input multi-output (MIMO) communication systems. Following a worst-case deterministic model, the actual channel is assumed to be inside an ellipsoid centered at a nominal channel. The objective is to maximize the worst-case received signal-to-noise ratio (SNR), or to minimize the worst-case Chernoff bound of the error probability, thus leading to a maximin problem. Moreover, we also consider the QoS problem, as a complement of the maximin design, which minimizes the transmit power consumption and meanwhile keeps the received SNR above a given threshold for any channel realization in the ellipsoid. It is shown that, for a general class of power constraints, both the maximin and QoS problems can be equivalently transformed into convex problems, or even further into semidefinite programs (SDPs), thus efficiently solvable by the numerical methods. The most interesting result is that the optimal transmit directions, i.e., the eigenvectors of the transmit covariance, are just the right singular vectors of the nominal channel under some mild conditions. This result leads to a channel-diagonalizing structure, as in the cases of perfect CSIT and statistical CSIT with mean or covariance feedback, and reduces the complicated matrix-valued problem to a scalar power allocation problem. Then we provide the closed-form solution to the resulting power allocation problem.

Published in:

IEEE Transactions on Signal Processing  (Volume:57 ,  Issue: 8 )