On the Performance of the Sum-of-Sinusoids Simulation Models for Rayleigh Fading Channels

This paper deals with the Sum-Of-Sinusoids (SOS)-based simulation method for generating timecorrelated Rayleigh fading processes. Five of the most accepted and recently proposed SOS-based Rayleigh fading models have been analyzed and compared. We have used the mean and the maximum power margin quality measures to assess the accuracy of the autocorrelation function (ACF) of the generated processes. The quality of the envelope probability density function (PDF) is evaluated using the Jeffrey's divergence (J-div) criterion. A comparison in terms of the time of execution is additionally presented. Compared with both stochastic and deterministic SOS models, the recently published Wang’s model has proven to be very successful. This model is shown to outperform all SOS models. Although Zheng and Xiao have recently introduced new SOS models, these latter have failed to achieve the performances reached by their first model which remains as the best choice among the Zheng and Xiao’s family of models. Key-Words: Clarke’s model; Rayleigh fading simulators; Sum-of-Sinusoids method; Statistical properties.