H. Du, P.-J. Chung, J. Gondzio and B. Mulgrew
Abstract
Transmit beamforming is a powerful technique for enhancing
performance of wireless communication systems. Most existing
transmit beamforming techniques require perfect channel
state information at the transmitter (CSIT), which is typically
not available in practice. In such situations, the design
should take into account errors in the channel estimates,
so that the beamformers are less sensitive to these errors.
Two robust approaches are widely used. The stochastic approach
optimizes the average performance of the system and
assumes that the statistics, such as mean and covariance, of
the errors are known. The maximin approach assumes that
the errors belong to a worst-case uncertainty region and optimizes
the worst-case system performance. This type of design
usually leads to conservative results as the worst-case
conditions may occur at a very low probability. In this paper,
we propose a more flexible approach that optimizes the average
beamforming performance and takes the extreme (but
rare) conditions into account proportionally. Simulation results
show that the proposed beamformer offers higher robustness
against errors in CSIT than serval state-of-the-art beamformers.
Key words: Wireless Communication Systems, Beamforming, Probabilistic Constraints, Robust Optimization.