Preserving a reduced feedback load in Multiuser Diversity systems is crucial. By establishing either single or multiple thresholds only in the energy domain, the Selective Multiuser Diversity (SMUD) schemes achieve a reduced feedback load while maintaining a high capacity performance. This paper introduces an enhanced joint time-energy domain scheduling approach for SMUD systems that maximizes the total spectral efficiency and preserves a minimal feedback load with just a few thresholds in the joint domain. This joint domain capitalizes on the energy and the non-fade duration statistics in order to establish a feedback policy that is focused on those users who have good channel condition and preserve it for a predetermined amount of time, providing thus an automatically more rational and fair use of the resources without a sophisticated scheduling with large polling/decision times. The analysis of this enhanced approach for the general case of Nakagami fading channels is developed and expressions for the average system capacity and the average feedback load are presented. The tradeoff between the feedback load and the spectral efficiency gain is analyzed. Two main scenarios are studied; the first when there are two thresholds in the energy domain side and one in the time domain side; the second when there are two thresholds in both domains.

The minimum duration outage concept allows an improved analysis of the outage probability and it has been reported, up to now, only for Log-normal shadowing and Rayleigh fading using asymptotic results of the level crossing theory. Capitalizing on some approximate results for the Fade Duration Distribution, FDD, in this letter we show the analysis of the minimum duration outage for the general case of Nakagami fading. We also show how some elements of the physical layer of the system can be used to determine the minimum duration outage interval ¿min.

The problem related to the fade and non-fade duration distributions (FDD and NFDD) of randomly distributed signals has been studied for more than forty years. However, for non-gaussian processes, closed form general solutions for these distributions are still unknown. In this paper we use an orthogonal series expansion to analyze the FDD and NFDD (also known as downcrossing and upcrossing duration distributions respectively) for the rather general case of Nakagami distributed random processes. Our results show that for asymptotic and non-asymptotic threshold levels the crossing duration distributions can be approximated, in a simple closed form way, using the gamma distribution. We back up this approximation with some goodness of fit metrics and show how the parameters of such distribution can be computed a priori analytically.

During almost fifty years the fade duration distribution (FDD) of random signals has been an open research topic. However, for Nakagami fading channels, and in general for the great majority of non-Gaussian processes, closed form general solutions are still unknown. In this paper we apply an orthogonal series expansion that allows to characterize the FDD. Our results show that the FDD in Nakagami wireless channels can be approximated, for both asymptotic and non-asymptotic threshold levels (C), by a simple closed form function: the Gamma distribution. This approximation is validated in different ways. We show how it is possible to compute (a priori) the parameters of the approximate Gamma distribution through the analytical fade statistics. We also examine the influence of different autocorrelation functions on the FDD.