Discretization methods of continuous time signals over frequency-flat fast-fading rayleigh channels

Discretization of the received continuous time signal is an essential stage in any receiver for a digital communication system. Over fast fading channels, inappropriate discretization of the received signal can result in SNR performance losses and high error floors. This paper presents a new framework for discrete representation of continuous time signals over frequency-flat time selective Rayleigh fading channels. This framework can serve as a basis for deriving receiver front-ends performing the task of received signal discretization. A key point of this framework is the representation of the zero-order Bessel function of the first kind in the form of a finite dimensional separable kernel. Two techniques that achieve this task are introduced, and the accuracy of these two approaches is investigated. Finally, a comparison of these two methods is provided.