Scribd
Upload
11 views5 pages

DnsLab Expt 2 SamplingTheorem

The document outlines an experiment to validate the Classical Sampling Theorem by designing a software simulation that explores the performance of sampling and reconstruction of bandlimited signals at various sampling rates. It details the theoretical background, including the Nyquist-Shannon Sampling Theorem, aliasing, and sinc interpolation, and specifies the expected outcomes, such as successful signal reconstruction and visualization of sampling effects. The experiment also includes practical steps for implementation using MATLAB, focusing on the impact of sampling rates and truncation of sinc functions on reconstruction accuracy.

Uploaded by

utkarshbabu.24
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
11 views5 pages

DnsLab Expt 2 SamplingTheorem

The document outlines an experiment to validate the Classical Sampling Theorem by designing a software simulation that explores the performance of sampling and reconstruction of bandlimited signals at various sampling rates. It details the theoretical background, including the Nyquist-Shannon Sampling Theorem, aliasing, and sinc interpolation, and specifies the expected outcomes, such as successful signal reconstruction and visualization of sampling effects. The experiment also includes practical steps for implementation using MATLAB, focusing on the impact of sampling rates and truncation of sinc functions on reconstruction accuracy.

Uploaded by

utkarshbabu.24
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

EXPERIMENT NO.

02 –The Classical Sampling Theorem

Group Number:
Name:
Roll Number:

____________________________________________________

Design and Simulation Laboratory-Communication Systems


Department of Electronics & Electrical Communication Engineering
Indian Institute of Technology Kharagpur
____________________________________________________
TITLE: The Classical Sampling Theorem
________________________________________________________________________________________

PROBLEM STATEMENT: To design a software experiment to validate the sampling theorem. Consider
the space of bandlimited functions, where signals are bandlimited to a maximum frequency of 30kHz. Design
a sampler and a corresponding interpolator, inspired by the classical Shannon-Nyquist-Whittaken Sampling
theory, and explore the performance of such a scheme (in terms of reconstruction error) for the following
sampling rate choice: (i) below the Nyquist rate (undersampling or sub-Nyquist sampling), (ii) at exactly the
Nyquist rate (optimal sampling or Nyquist rate sampling), (iii) above Nyquist rate (super-Nyquist sampling).
Generate visual plots (both in time and frequency domain) for the original signal (from the bandlimited class
of signals), the sampled signal and the reconstructed signal.

_________________________________________________________________________________

KEY COMPONENTS: PC with Matlab

<Add more as necessary>

BRIEF THEORY & BACKGROUND:

The experiment aims to generate a bandlimited signal perform uniform sampling operation and reconstruct
back the original signal using sinc (cardinal sine) interpolation.

Sampling is the process of converting a continuous-time signal into a discrete-time signal by taking
measurements at uniform time intervals. This is an essential step in digital signal processing, enabling the
analysis and processing of signals using digital systems. Sampling is used in various applications, including
telecommunications, audio and video processing, and control systems. The sampling process enables the
representation and manipulation of signals in a digital form, which is crucial for storage, transmission, and
processing using digital systems.

SAMPLING PROCESS

In the sampling process, a continuous signal x(t) is converted into a discrete set of values x[n], where n is an
integer. The relationship between the continuous-time signal and the discrete-time signal is defined by:

x[n]=x(nTs)

where Ts is the sampling interval, and fs:=1/Ts is the sampling frequency. The sampling frequency determines
how many samples are taken per second.

Nyquist-Shannon Sampling Theorem:


The Nyquist-Shannon Sampling Theorem provides the criteria for perfect signal reconstruction. Essentially, it
states that a band-limited signal with a maximum frequency fm can be perfectly reconstructed from its
samples if the sampling frequency fs is greater than or equal to twice the maximum frequency:

fs>=2fm

This condition is known as the Nyquist rate. Sampling at or above this rate ensures that no information is lost
during the sampling process, and the original continuous-time signal can be completely recovered.

Aliasing

Aliasing occurs when a signal is sampled below its Nyquist rate. In this situation, higher frequency
components of the signal are indistinguishably mapped to lower frequencies, leading to distortion in the
reconstructed signal. To prevent aliasing, signals should be filtered with an anti-aliasing filter before
sampling. This filter removes frequency components above fm to ensure that only the desired frequency
range is sampled.

Sinc Interpolation

If the sampling rate is sufficiently high, performing interpolation using appropriate sinc (cardinal sine)
function reconstructs a continuous signal from its sampled data. In particular, each discrete sample is multiplied
by a sinc function centered at that sample, and the sum of these weighted sinc functions forms the continuous
signal, say xr(t) which equals the original bandlimited function x(t), i.e.,

This method provides perfect reconstruction for band-limited signals, assuming an ideal sampling process
and infinite sinc functions.

Truncated Sinc Interpolation

In practical applications, sinc functions are truncated to a finite length due to computational limitations. A
truncated sinc function limits the range of the sinc function to a specific interval, say for instance an interval
[−kT,kT] , where T=1/fs. Although truncation introduces some error, it reduces computational complexity
and makes real-time processing feasible.

<Add a formal state of the sampling theorem and a proof of the sampling theorem>
EXPECTED OUTCOMES:

1. Successful Signal Reconstruction:

The interpolated signal should closely match the original continuous-time signal when sampled at a rate above
the Nyquist frequency.

2. Visualization of Sampling:

Clear plots showing the continuous signal, discrete sampled points, and the reconstructed signal.

3. Aliasing Demonstration:

If the sampling rate is below the Nyquist rate, the reconstructed signal should show distortion or aliasing effects.

4. Impact of Truncated Sinc Function:

The truncated sinc function should allow for efficient signal reconstruction, with minimal artifacts due to
truncation.

PROBLEM SETUP AND EXPERIMENT DETAILS:

Assume the following for the experiment:


1. Bandwidth of the signal(fm): 30kHz.
1. Take a (discrete) time vector from 0 to T, where T can be any particular time value (large enough).
2. Set sampling frequency fs, and set signal frequency fm.
3. Generate a continuous time bandlimited signal, say signal_cont or x(t) and plot it over the T instants.
4. Perform sampling operation with sampling period 1/fs and plot the sampled_signal, say xsp(t)
5. Use inbuilt sinc function to perform sinc interpolation based on the sampled data. Let us denote the
reconstructed signal as reconstructed_signal or xr(t).
6. Now plot the signal_cont (x(t)) and reconstructed_signal (xr(t)) and capture the reconstruction error, say
e(t). Plot it. Comment on it.
7. Present a Fourier-domain or spectral domain set of plots and comment on them.
8. Perform this same operation for different sampling rates i.e, undersampled and oversampled case and
check rconstruction is possible or not, plot the graphs and comment on those plots similarly.
9. Since sinc pulses decay only at infinite time, let us study the effect of truncation: instead of taking the
whole sinc pulse, truncate the sinc pulse to main lobe, i.e., let us fix the window as [-Ts,Ts] (where Ts is the
reciprocal of fs) and use the new truncated sinc pulse to perform the interpolation across the samples
generated. Repeat the experiment where the sinc pulse is truncated to two side lobes, i.e.,[-2T s ,2Ts] and
perform same above operation. Similarly perform truncation operation to 3 lobes, 5 lobes, and 7 lobes and
repeat above operation. Present the plots for the original function x(t), and its reconstructions under each of
the settings above (color code them and clearly mention in the legend).
10. Now calculate the error for each truncation case and plot the graphs of no of lobes truncated in sinc vs the
error.

EXTRA CREDIT:

Using the theory of sampling, setup a MATLAB experiment where you capture audio data using the MIC and
then sample and interpolate (with appropriate sampling rate as per the bandlimit for voice signals). Play the
reconstructed signal to validate the “goodness” of the sampling-reconstruction scheme.

ACKNOWLEDGEMENTS:

Mention (enumerate) all references including private communication, reference papers, public articles, etc.
Look up online resources to understand how references are cited.

REFERENCES:

[1]

<Add appropriate references>

You might also like