Thèse de doctorat en Lasers, métrologie et communications
Soutenue en 2009
à Paris, CNAM .
Pas de résumé disponible.
Pas de résumé disponible.
Conribution to multi-antenna Multi-user downlink communication systems with feedback
The next generation of wireless networks will require more efficient use of the system resources to accommodate the demand for voice, video, and data transmission. In this thesis, we investigate methods to increase the transmission rate in three different downlink multi-antenna systems : the first and the second are respectively the single user and multi-user systems with partial channel state information (CSIT), the third one is the multicast multi-user system. For the first two systems, we studied and compared two feedback methods : the limited feedback, where a quantized version of the CSI is transmitted to the scheduler and the analog feedback method. For the multi-antenna multi-user systems,we proposed a technique to reduce the feedback load for multi-antenna multi-user system,this technique can at the same time preserve the multi-user diversity while reducing the complexity of the user selection algorithm. A part of the proposed user-selection algorithm is implemented on the users’ sides. The feedback load on the uplink channels is controlled by thresholds which are determined by the base station according to the number of users in the cell. We showed that in this case the system performances are more affected by the size quantization of the channel direction than the size quantization of the channel magnitude. We also derived a lower bound on the ergodic sum rate for the proposed system. We also studied the single and multi-user system performance when the feedback channels are affected by the noise. In the case of the multi-user system, we proposed an algorithm to reduce this noise on the feedback channels. Finally, in the case of the multi-user muticast system, we proposed two user selection algorithms in order to maximize the total sumrate. We analyze the performance of these algorithms in multicast orthogonal frequency division multiplexing (OFDM) systems with single and multiple transmit antennas. We showed that the resource allocation that includes the subcarrier allocation, bit loading, and the precoding vector selection in the multiple-input single-output (MISO) case is a difficult optimization problem. Consequently, we propose suboptimal algorithms based on the maximization of the sum data rate and the maximization of the minimum user data rate criteria.