Performance Analysis and Optimization of Best-M Feedback for OFDMA Systems

In this paper, the best-M feedback, where each user reports the channel quality indicators (CQIs) on its M best resource blocks (RBs), for orthogonal frequency division multiple access (OFDMA) systems is analyzed and optimized. First, the closed-form expression of the average sum-rate is derived as a function of M and the number of bits for a signal to noise ratio (SNR) quantization B. Then, M and B are jointly optimized to minimize the feedback overhead such that the desired system performance can be achieved. Numerical results confirm that the proposed analysis is quite well matched to the exact result so that our work is useful to optimize M and B instead of time-consuming computer simulations, and that the required feedback overhead to achieve a given performance decreases as the number of users and the average SNR increase.