A Channel Quality Metric in Opportunistic Selection With Outdated CSI Over Nakagami- $m$ Fading Channels

In opportunistic selection (OS)-based systems, outdated channel state information (CSI) is encountered due to channel variation between the instants of channel estimation and actual data transmission. This will result in unreliable CSIs employed in OS, which greatly degrades the system performance. To address this issue, a novel metric is proposed to evaluate the transmission quality of a channel suffering from Nakagami-
$m$
fading, taking both outdated CSI and fading statistics into consideration. Moreover, its suboptimized counterpart is also proposed to significantly reduce computational complexity. We then use this metric to select users (sources) in a multiuser scheduling model and evaluate its outage performance. Both analytical and simulation results show that our proposals outperform the existing alternatives in terms of outage probability.