ELECTRONICS & COMMUNICATION ENGINEERING

DIGITAL SIGNAL PROCESSING CATEGORY L T P CREDIT

ECL333
LABORATORY PCC 0 0 3 2

Preamble:
The following experiments are designed to make the student do real time DSP
• computing.

Dedicated DSP hardware (such as TI or Analog Devices development/evaluation

•
boards) will be used for realization.

Prerequisites:
• ECT 303 Digital Signal Processing

• EST 102 Programming in C

Course Outcomes: The student will be able to

CO 1 Simulate digital signals.
CO 2 verify the properties of DFT computationally
CO 3 Familiarize the DSP hardware and interface with computer.
CO 4 Implement LTI systems with linear convolution.
CO 5 Implement FFT and IFFT and use it on real time signals.
CO 6 Implement FIR low pass filter.
CO 7 Implement real time LTI systems with block convolution and FFT.

Mapping of Course Outcomes with Program Outcomes

PO PO PO PO PO PO PO PO PO PO1 PO1 PO1
1 2 3 4 5 6 7 8 9 0 1 2
CO1 3 3 1 2 3 0 0 0 3 0 0 1
CO2 3 3 1 2 3 0 0 0 3 0 0 1
CO3 3 3 3 2 3 0 0 0 3 1 0 1
CO4 3 3 1 2 3 0 0 0 3 0 0 1
CO5 3 3 1 1 3 0 0 0 0 0 0 1
CO6 3 3 1 1 3 0 0 0 0 0 0 1
CO7 3 3 1 3 3 0 0 0 3 3 0 0
 ELECTRONICS & COMMUNICATION ENGINEERING

Assessment Pattern

Mark Distribution:

Total Mark CIE ESE

150 50 100

Continuous Internal Evaluation Pattern:

Each experiment will be evaluated out of 50 credits continuously as

Attribute Mark
Attendance 15
Continuous assessment 30
Internal Test (Immediately before 30
the second series test)

End Semester Examination Pattern: The following guidelines should be followed

regarding award of marks

Attribute Mark
Preliminary work 15
Implementing the work/ 10
Conducting the experiment
Performance, result and inference 25
(usage of equipments and trouble
shooting)
Viva voce 20
Record 5

Course Level Assessment Questions

CO1-Simulation of Signals

1. Write a Python/MATLAB/SCILAB function to generate a rectangular

pulse.
2. Write a Python/MATLAB/SCILAB function to generate a triangular
pulse.

CO2-Verfication of the Properties of DFT

1. Write a Python/MATLAB/SCILAB function to compute the N -point DFT

 ELECTRONICS & COMMUNICATION ENGINEERING

matrix and plot its real and imaginary parts.

2. Write a Python/MATLAB/SCILAB function to verify Parseval’s theorem

for N = 1024.

CO3-Familarization of DSP Hardware

1. Write a C function to control the output LEDs with input switches.

2. Write a C function to connect the analog input port to the output port and test with

a microphone.

CO4-LTI System with Linear Convolution

1. Write a function to compute the linear convolution and download to the hardware
target and test with some signals.

CO5-FFT Computation

1. Write and download a function to compute N point FFT to the DSP hardware
target and test it on real time signal.
2. Write a C function to compute IFFT with FFT function and test in on DSP
hardware.

CO6-Implementation of FIR Filter

1. Design and implement an FIR low pass filter for a cut off frequency of 0.1π and
test it with an AF signal generator.

CO7-LTI Systems by Block Convolution

1. Implement an overlap add block convolution for speech signals on DSP

target.
 ELECTRONICS & COMMUNICATION ENGINEERING

List of Experiments
(All experiments are mandatory.)

Experiment 1. Simulation of Signals Simulate the following signals using Python/

Scilab/MATLAB.
1. Unit impulse signal
2. Unit pulse signal
3. Unit ramp signal
4. Bipolar pulse
5. Triangular signal

Experiment 2. Verification of the Properties of DFT

• Generate and appreciate a DFT matrix.

1. Write a function that returns the N point DFT matrix VN for a given
N.
2. Plot its real and imaginary parts of VN as images using matshow or
imshow commands (in Python) for N = 16, N = 64 and N = 1024
3. Compute the DFTs of 16 point, 64 point and 1024 point random
sequences using the above matrices.
4. Observe the time of computations for N = 2γ for 2 γ 18≤(You
≤ may use
the time module in Python).
5. Use some iterations to plot the times of computation against γ. Plot
and understand this curve. Plot the times of computation for the fft
function over this curve and appreciate the computational saving
with FFT.

• Circular Convolution.
1. Write a python function circcon.py that returns the circular con-
voluton of an N1 point sequence and an N2 point sequence given at
the input. The easiest way is to convert a linear convolution into
circular convolution with N = max(N1, N2).

• Parseval’s Theorem
For the complex random sequences x1[n] and x2[n],

N −1 N −1
X 1 X
x1 [n]x∗2 [n] = X1 [k]X2∗ [k]
n=0
N k=0
 ELECTRONICS & COMMUNICATION ENGINEERING

1. Generate two random complex sequences of say 5000 values.

2. Prove the theorem for these signals.

Experiment 3. Familarization of DSP Hardware

1. Familiarization of the code composer studio (in the case of TI hard- ware)
or Visual DSP (in the case of Analog Devices hardware) or any equivalent
cross compiler for DSP programming.
2. Familiarization of the analog and digital input and output ports of the DSP
board.
3. Generation and cross compilation and execution of the C code to con- nect
the input digital switches to the output LEDs.
4. Generation and cross compilation and execution of the C code to con- nect
the input analog port to the output. Connect a microphone, speak into it
and observe the output electrical signal on a DSO and store it.
5. Document the work.

Experiment 4. Linear convolution

1. Write a C function for the linear convolution of two arrays.

2. The arrays may be kept in different files and downloaded to the DSP
hardware.
3. Store the result as a file and observe the output.

4. Document the work.

Experiment 5. FFT of signals

1. Write a C function for N - point FFT.

2. Connect a precision signal generator and apply 1 mV , 1 kHz sinusoid at

the analog port.

3. Apply the FFT on the input signal with appropriate window size and
observe the result.
4. Connect microphone to the analog port and read in real time speech.

5. Observe and store the FFT values.

6. Document the work.

 ELECTRONICS & COMMUNICATION ENGINEERING

Experiment 6. IFFT with FFT

1. Use the FFT function in the previous experiment to compute the IFFT of
the input signal.
2. Apply IFFT on the stored FFT values from the previous experiments and
observe the reconstruction.
3. Document the work.

Experiment 7. FIR low pass filter

1. sin(ω
Use Python/scilab to implement the FIR filter response h[n] = cn)
πn
for a filter size N = 50, ωc = 0.1π and ωc = 0.3π .

2. Realize the hamming(wH [n]) and kaiser (wK[n]) windows.

3. Compute h[n]w[n] in both cases and store as file.

4. Observe the low pass response in the simulator.

5. Download the filter on to the DSP target board and test with 1 mV
sinusoid from a signal generator connected to the analog port.

6. Test the operation of the filters with speech signals.

7. Document the work.

Experiment 8. Overlap Save Block Convolution

1. Use the file of filter coefficients From the previos experiment.

2. Realize the system shown below for the input speech signal x[n].

3. Segment the signal values into blocks of length N = 2000. Pad the last
 ELECTRONICS & COMMUNICATION ENGINEERING

block with zeros, if necessary.

4. Implement the overlap save block convolution method

5. Document the work.

Experiment 9. Overlap Add Block Convolution

1. Use the file of filter coefficients from the previous experiment.

2. Realize the system shown in the previous experiment for the input speech
signal x[n].

3. Segment the signal values into blocks of length N = 2000. Pad the last
block with zeros, if necessary.

4. Implement the overlap add block convolution method

5. Document the work.

Schedule of Experiments: Every experiment should be completed in three hours.

Textbooks

1. Vinay K. Ingle, John G. Proakis, “Digital Signal Processing Using

MATLAB.”

2. Allen B. Downey, “Think DSP: Digital Signal Processing using Python.”

3. Rulph Chassaing, “DSP Applications Using C and the TMS320C6x DSK

(Topics in Digital Signal Processing)”