JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech III Year DIGITAL COMMUNICATIONS Previous Question Paper (July/August - 2023 - R16)

Flat-top sampling is a practical method of sampling in which the top of each sample pulse is maintained at a constant amplitude, equal to the instantaneous value of the continuous-time signal at the sampling instant. Unlike ideal impulse sampling, where the sample is a Dirac delta function, flat-top sampling produces pulses with a finite duration.

In this process, the input signal x(t) is sampled at intervals of T_s, and the resulting pulse is held constant for the duration of the pulse width τ. This technique is commonly used in Pulse Amplitude Modulation (PAM) systems because it simplifies the recovery process and reduces the complexity of the hardware required for subsequent processing.

Mathematically, the sampled signal x_s(t) can be represented as the convolution of the ideal sampled signal with a rectangular pulse h(t):

where h(t) is a rectangular pulse of width τ. A significant effect of flat-top sampling is the aperture effect, which introduces a distortion in the frequency spectrum of the sampled signal, characterized by a sinc-shaped attenuation, requiring a compensation filter at the receiver.

Delta Modulation (DM) is a simple form of analog-to-digital conversion, but it suffers from two primary limitations that restrict its performance in high-fidelity signal reconstruction:

• Slope Overload Distortion: This occurs when the input analog signal changes at a rate faster than the staircase approximation can follow. Because the step size is fixed, the system cannot keep up with rapid variations in the signal amplitude. Mathematically, this happens when the slope of the input signal exceeds the maximum slope of the DM system, defined as:
  Δ × fₛ > d/dt [m(t)]
  where Δ is the step size and fₛ is the sampling frequency.
• Granular Noise: This is the quantization error that occurs when the input signal is relatively flat or changes very slowly. In this state, the staircase output oscillates around the input signal with a constant step size, creating a noise component known as granular noise. This noise is proportional to the step size Δ, meaning that reducing Δ to minimize granular noise inherently exacerbates slope overload distortion.

Note: These two drawbacks are fundamentally linked; increasing the step size improves slope overload performance but increases granular noise, while decreasing the step size reduces granular noise but increases slope overload distortion.