UC San Diego

Multiuser diversity gain usually increases with the number of independent users in a system and with the dynamic range of the channel fluctuation. To enhance the sum rate of a system, joint consideration of scheduling and traditional diversity schemes such as selection diversity, combining diversity, and coded diversity is addressed in past works. The basic principle of joint consideration is to enhance multiuser diversity gain by increasing the number of independent candidates for selection in direct proportion to the number of transmit or receive antennas, or by increasing the variation in the channels between the transmitter and receivers. In the first part of the dissertation, we consider joint scheduling and diversity with low feedback requirements to enhance the benefits of multiuser diversity in an orthogonal frequency division multiple access (OFDMA) system. The OFDMA spectrum is assumed to consist of N_RB resource blocks and the reduced feedback scheme consistsof each user feeding back channel quality information (CQI) for only the best-N_FB resource blocks. Assuming largest normalized CQI scheduling and a general value for N_FB, we develop a unified framework to analyze the sum rate of the system for both the quantized and non-quantized CQI feedback schemes. Based on this framework, we provide closed-form expressions for the sum rate for three di®erent multi-antenna transmitter schemes; Transmit antenna selection, orthogonal space time block codes and cyclic delay diversity (CDD). Furthermore, we approximate the sum rate expression and determine the feedback ratio (N_FB/N_FB) required to achieve a sum rate comparable to the sum rate obtained by a full feedback scheme. In the second part of the dissertation, we examine the interplay between frequency selectivity and multiuser diversity in an OFDMA scheduling system. A scheduling unit block consists of contiguous multiple subcarriers. Users are scheduled based on their block average throughput in a proportional fair way. Multiuser diversity gain increases with the degree and dynamic range of channel °uctuations. However, a decrease of the block average throughput in a highly selective channel can lower the sum rate as well. In this part, we first study channel selectivity that is desirable for maximizing the maximum of the block average throughput of an arbitrary user. Based on this study, we then propose a method to determine per-user optimal cyclic delay when CDD is used to enhance the sum rate by increasing channel selectivity for a channel with limited frequency selectivity. We show that the throughput by the proposed technique is very close to the optimal sum rate possible with CDD. In the third part of the dissertation, we consider joint user scheduling and beamforming to enhance sum-rate performance in an interference-limited uplink cellular network, and propose three schemes for this purpose. Specifically, one method is to maximize the signal to generated interference and noise ratio and another is to maximize the signal to interference and noise ratio in the dual network. To improve the user- orthogonality between selected users, we also propose two- step user-selection procedure. We compare the proposed schemes with a conventional scheme which maximizes the signal to noise ratio or minimizes generated interference to other cells. We show that the proposed schemes outperform the conventional ones in most cases, and better exploit multiuser diversity in reducing inter-cell interference