Constrained power allocation algorithm for rate adaptive MIMO system

We study the power allocation problem in wireless multiple-input multiple-output (MIMO) systems adopting QAM modulation. In addition to the total power constraint, two additional practical constraints are included in our study. Firstly, the quality of service (QoS) specification is expressed in terms of required BER for each subchannel. Secondly, only a few discrete constellation sizes are available for signal modulation. We analyze the power allocation problem for such systems, and introduce a new performance index, so-called throughput-improvement-to-power-increment ratio (PITIR), to reveal the underlying idea of the strategy. We then propose a power allocation algorithm for maximizing the data throughput given total power budget. The proposed algorithm is simple and efficient for implementation, and it achieves almost the same data throughput performance and total power consumption as a previously published scheme. In addition, the proposed algorithm can make use of computational results obtained in earlier transmissions, and this reduces the overall computation load.