Scribd
Upload
24 views

DSP 1

Uploaded by

naveensl0207
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
24 views

DSP 1

Uploaded by

naveensl0207
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Signal Sampling and Quantization

by Dr. W. G. Chathuranga Kumarage

1 Introduction
This lecture explores the fundamental processes of **Signal Sampling** and
**Quantization**, which are critical in converting continuous-time signals to
digital format. We will also discuss **signal reconstruction** and practical
concerns like anti-aliasing filtering.

2 Signal Sampling
Sampling is the process of converting a continuous-time signal into a discrete-
time signal by taking samples at regular intervals.

2.1 Sampling Process


Sampling can be mathematically described as:
∞
X
xs (t) = x(nTs )δ(t − nTs )
n=−∞

where Ts is the sampling period, and x(nTs ) is the sampled value at time t =
nTs .

1
2.2 Schematic: Sampling Process

Amplitude
xs (t)

Time

Sampled points

Figure 1: Illustration of the sampling process, where a continuous signal is


sampled at regular intervals.

2.3 Nyquist-Shannon Sampling Theorem


To prevent information loss (aliasing), the **Nyquist-Shannon sampling theo-
rem** states that the sampling frequency fs must be at least twice the highest
frequency component of the signal, i.e.,

fs ≥ 2fmax

2.4 Aliasing
When a signal is sampled below the Nyquist rate, it causes overlapping of fre-
quency components, leading to **aliasing**.

2.5 Solved Example: Audio Signal Sampling


Problem: An audio signal has a maximum frequency of 15 kHz. What is the
minimum sampling frequency required?
Solution: The minimum sampling frequency fs must be twice the highest
frequency:
fs = 2 × 15 kHz = 30 kHz
Thus, the signal must be sampled at 30 kHz or higher.

3 Quantization
Quantization maps continuous amplitude values of a sampled signal into discrete
values, introducing **quantization error**.

2
3.1 Quantization Process
The quantization process can be defined as follows:
1. **Sampling the Signal:** The continuous signal is first sampled to obtain
discrete amplitude values.

2. **Determining the Number of Levels:** The number of quantization lev-


els, L, is determined based on the bit depth b:

L = 2b

where b is the number of bits used for quantization.


3. **Calculating Step Size:** The quantization step size ∆ can be calculated
as:
Vmax − Vmin
∆=
L
where Vmax and Vmin are the maximum and minimum values of the signal,
respectively.
4. **Rounding to Nearest Level:** Each sampled value x(nTs ) is then rounded
to the nearest quantization level. This can be mathematically represented
as:
xq (nTs ) = Q(x(nTs ))
where Q(·) is the quantization function.
5. **Output the Quantized Signal:** The resulting quantized signal will have
values corresponding to the quantization levels.

3.2 Schematic: Quantization Levels

Amplitude

Quantized
Quantized
Level Level
Quantized Value
Quantized Level Quantized Level Continuous
Time Signal
Sampled Value
Quantized
Quantized
Level Level

Quantization Level
SampledSampled
Point Sampled
Point Sampled
Point Sampled
Point Sampled
Point Point

Figure 2: Mapping of a continuous signal to discrete quantization levels. Sam-


pled values are rounded to the nearest quantization level.

3
3.3 Quantization Error
Quantization introduces an error defined as the difference between the actual
sampled value and the quantized value. The maximum quantization error Emax
is given by:
∆
Emax =
2
where ∆ is the quantization step size.

3.4 Schematic: Quantization Error

Amplitude

Quantized Value
Quantization Error Value
Original Sampled Point

Figure 3: Quantization error represented as the difference between the actual


and quantized values.

3.5 Practical Considerations


1. **Bit Depth Impact:** Increasing the bit depth improves the resolution of
the quantized signal, resulting in lower quantization noise and higher signal
fidelity.
2. **Trade-offs:** While increasing bit depth improves quality, it also in-
creases the amount of data that needs to be processed and stored.
3. **Non-linear Quantization:** In some applications, non-linear quanti-
zation techniques (e.g., companding) may be used to allocate more levels to
smaller amplitude signals, which are more perceptually significant.

3.6 Solved Example: Image Signal Quantization


Problem: An image is quantized using 8-bit gray levels (256 levels). If the
number of levels is increased to 1024 (10-bit), how much improvement in SQNR
can be expected?
Solution: Using the approximation for SQNR:
SQNR improvement = 6 × (difference in bits) = 6 × (10 − 8) = 12 dB
Increasing by 2 bits improves SQNR by 12 dB.

4
4 Signal Reconstruction
Signal reconstruction is the process of converting a discrete-time signal back
into a continuous-time signal. This involves using a **reconstruction filter**
that smooths out the discrete samples to form a continuous signal.

4.1 Steps in Signal Reconstruction


The reconstruction process typically involves the following steps:
1. **Sampling the Signal:** The original continuous signal is sampled at
intervals to produce discrete samples.
2. **Storage or Transmission:** The discrete samples are either stored or
transmitted for later use.
3. **Reconstruction Filtering:** A reconstruction filter (often a low-pass fil-
ter) is applied to the sampled signal to smooth out the transitions between
samples. This filter should ideally have a cutoff frequency just above the
highest frequency of the original signal.

4. **Output of Reconstructed Signal:** The output of the reconstruction


filter is a continuous-time signal that approximates the original signal.

4.2 Mathematical Description


The reconstructed signal xr (t) can be described mathematically by convolving
the discrete-time signal with a sinc function:
∞  
X t − nTs
xr (t) = x(nTs ) · sinc
n=−∞
Ts

sin(πt)
where sinc(t) = πt serves as the ideal low-pass filter.

5
4.3 Schematic: Signal Reconstruction Process

Amplitude
Reconstructed signal

Time

Figure 4: Signal reconstruction from sampled values. The reconstruction filter


generates a smooth continuous signal.

4.4 Practical Considerations


- **Reconstruction Filters:** The choice of reconstruction filter can significantly
affect the quality of the reconstructed signal. Ideal filters are difficult to imple-
ment, so practical filters are used, which may introduce some distortion.
- **Anti-Aliasing:** To avoid aliasing during sampling, an anti-aliasing filter
is often applied to the signal before sampling. This filter removes frequency
components above the Nyquist frequency.

5 Exercises
• Exercise 1: A signal with a bandwidth of 10 kHz is sampled. What is
the minimum sampling frequency to avoid aliasing?
• Exercise 2: An 8-bit quantizer is used for a signal ranging from -5V to 5V.
What is the quantization step size? What is the maximum quantization
error?
• Exercise 3: Explain the effect of increasing the bit depth from 8-bit to
16-bit on quantization noise and signal accuracy.

6 Further Reading
To explore these concepts further, consider the following resources:
• Digital Signal Processing: Principles, Algorithms, and Applications by
Proakis and Manolakis.
• Practical examples using MATLAB for signal processing.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy