Mode switching for the multi-antenna broadcast channel based on delay and channel quantization
EURASIP Journal on Advances in Signal Processing - Multiuser MIMO Transmission with Limited Feedback, Cooperation, and Coordination
Cross-Layer Antenna Beamforming and Power Control in Wireless Uplinks
Wireless Personal Communications: An International Journal
GLOBECOM'09 Proceedings of the 28th IEEE conference on Global telecommunications
Degrees of freedom and sum rate maximization for two mutually interfering broadcast channels
ICC'09 Proceedings of the 2009 IEEE international conference on Communications
Fairness-aware radio resource management in downlink OFDMA cellular relay networks
IEEE Transactions on Wireless Communications
Multiuser MIMO achievable rates with downlink training and channel state
IEEE Transactions on Information Theory
Limited feedback for temporally correlated MIMO channels with other cell interference
IEEE Transactions on Signal Processing
Spatial multiplexing gain for two interfering MIMO broadcast channels based on linear transceiver
IEEE Transactions on Wireless Communications
Distributed beamforming techniques for weighted sum-rate maximization in MISO interference channels
IEEE Communications Letters
| Hi-index | 0.14 |
We consider the downlink of a wireless system where the base-station has M ges 1 antennas and K user terminals have one antenna each. We study the weighted rate sum maximization in the case of non-perfect Channel State Information at the Transmitter (CSIT). Some relevant downlink optimization problems, such as the stabilization of the transmission queues under random packet arrivals and the proportional fair scheduling for infinite backlogged systems, can be solved as special cases of the proposed problem. We restrict the transmitter strategy to be based on Gaussian coding and beamforming. Even under this simplifying condition, the problem at hand is non-convex and it does not appear to lend itself to a simple algorithmic solution. Therefore, we introduce some approximations that yield a definition of signal-to-interference plus noise ratio (SINR) commonly used in the classical array- processing/beamforming literature. For the simpler (but still non-convex) approximated problem, we propose a powerful heuristic solution based on greedy user selection and a gradient iteration that converges to a local maximum of the objective function. This method yields very competitive results with relatively low computational complexity. Extensive simulations show that, in the case of perfect CSIT, the proposed heuristic scheme performs very closely to the optimal (dirty-paper coding) strategy while, in the case of non-perfect CSIT, it significantly outperforms previously proposed suboptimal approaches, such as random beamforming and approximated zero-forcing with greedy user selection.