Multiuser diversity with quantized feedback

In this paper, we propose an optimal discrete rate switch-based multiuser diversity (DSMUDiv) scheduling scheme that reduces the feedback load while preserving most of the performance of opportunistic scheduling. In order to reduce the feedback rate, quantized values indicating the modulation level are fedback instead of the full values of the SNRs (quantized SNRs). We examine the DSMUDiv scheme using two scheduling criteria depending on the distribution of the mobile users in the cell: (i) absolute signal-to-noise ratio (SNR)-based scheduling in the case of independent and identical distributed (i.i.d.) users, and (ii) normalized SNR-based scheduling in the case of i.-non-i.d. users. The paper includes the derivation of closed-form expressions of the feedback load, spectral efficiency and probability of access. Monte Carlo simulation is used to evaluate the spectral efficiency in the case of normalized SNR-based scheduling. We compare our scheduling scheme under slow Rayleigh fading assumption with the optimal (full feedback load) selective diversity scheduling scheme.