Performance-guaranteed resource allocation in wireless communication systems

Growing user demands in current fourth generation (4G) and future fifth generation (5G) wireless communication systems necessitate efficient resource utilization. This work is motivated by the optimization potential of 4G systems employing orthogonal frequency division multiple access (OFDMA) technology. It covers centralized single-cell instantaneous resource allocation problems with perfect channel state information and adaptive modulation. The exemplary problems analyzed in this work are weighted sum rate maximization, weighted sum power minimization, fairness-constrained rate maximization, and heterogeneous rate maximization. We develop a general framework for resource allocation problems, which encompasses the common problem formulations found in the literature. We show that so-called rounding methods provide an approximation scheme based on linear relaxation. A general bound on the absolute performance loss of such methods is derived. This result yields a performance guarantee for the exemplary problems analyzed in this work. Under certain regularity conditions, the relative performance of rounding methods is shown to be asymptotically optimal as the number of subcarriers grows.