Dynamic subchannel and bit allocation for multiuser-MIMO/OFDMA systems

Abstract

In this thesis, adaptive subchannel and bit allocation schemes, which minimize the overall transmission power under a rate constraint, are proposed for various orthogonal frequency division multiple access (OFDMA) systems. First, for single-input-single-output (SISO)/OFDMA systems, an adaptive subcarrier and bit allocation scheme, which performs subcarrier allocation and bit loading separately, is developed. Specifically, based on the observation that the optimal approach tends to assign best subcarriers to a user, subcarrier allocation is performed. The bit loading, which follows the subcarrier allocation, can be performed in a manner similar to single-user OFDM. Computer simulation shows that the proposed suboptimal approach outperforms the existing ones. Second, an adaptive subchannel allocation and modulation scheme is developed for multiuser (MU)-multiple-input multiple-output (MIMO)/OFDMA systems through certain modifications of the approach for SISO/OFDMA systems. It is assumed that an MU-MIMO/OFDMA system employs singular value decomposition (SVD)-based SDMA and MIMO processors, which decompose an MU-MIMO channel into parallel single-user (SU)-MIMO channels consisting of SISO subchannels. Optimal and suboptimal procedures for allocating the SU-MIMO channels to users and loading bits to the SISO subchannels are proposed. It is shown that the optimization problem, which minimizes the overall transmission power under a rate constraint, is formulated as an integer programming (IP) problem. Then a suboptimal algorithm that separately performs subchannel allocation and bit loading is proposed. In particular, efficient algorithms are developed to solve the separate subchannel allocation problem based on modifying heuristic algorithms for transportation problems. Computer simulation results indicate that the performances of the suboptimal algorithms are reasonably close to that of the optimal IP, yet the former is considerably sim...