Further results on the asymptotic analysis of digital communications in generalized fading channels

Abstract

This articles presents several new and simple closed-form asymptotic approximations for the average symbol error rate (ASER) performance measure of digital communication systems (with/without diversity transceivers) impaired by additive white Gaussian noise (AWGN) and multipath fading. In particular, a new class of asymptotic approximations for the ASER is derived by utilizing a normalized asymptotic probability density function (PDF) of signal-to-noise ratio (SNR). The resulting expressions are considerably better/tighter over a wider range of mean SNRs when compared to the existing high-SNR approximations although both techniques need only the first non-zero term of the Maclaurin (if exists) or the Poincare series expansion of the receiver output SNR. We also investigate the utility/efficacy of the Welch-Satterthwaite and Moschopoulos approximations (based on either a moments matching technique or using our proposed normalized asymptotic PDF of SNR that resembles a Gamma PDF) for yielding accurate predictions of the ASER in the low-SNR regime in a myriad of fading environments. Simple closed-form approximations for the ergodic channel capacities of different types of fading channels with/without diversity reception are also derived.