Scribd
Upload
2 views

Tutorial-1 (1)

The document is a tutorial for ECC 511 - Digital Signal Processing & Digital Communication at IIT Roorkee, containing various problems related to discrete-time sequences, continuous-time signals, sampling rates, and Fourier transforms. It covers topics such as Nyquist rates, DTFT, DFT, and convolution, providing exercises for students to solve. The tutorial aims to enhance understanding of digital signal processing concepts through practical applications.

Uploaded by

bojo2backup
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Tutorial-1 (1)

The document is a tutorial for ECC 511 - Digital Signal Processing & Digital Communication at IIT Roorkee, containing various problems related to discrete-time sequences, continuous-time signals, sampling rates, and Fourier transforms. It covers topics such as Nyquist rates, DTFT, DFT, and convolution, providing exercises for students to solve. The tutorial aims to enhance understanding of digital signal processing concepts through practical applications.

Uploaded by

bojo2backup
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Department of Electronics & Communication Engineering

Indian Institute of Technology Roorkee


ECC 511 – Digital Signal Processing & Digital Communication
tutorial #1

1. Consider the discrete-time sequence


𝑛𝜋
𝑥 [𝑛] = cos ( )
8
Find two different continuous-time signals that would produce this sequence when sampled at a
frequency of 𝑓𝑠 = 10 kHz.

2. If the Nyquist rate for 𝑥𝑎 (𝑡) is Ω𝑠 , what is the Nyquist rate for each of the following signals that are
derived from 𝑥𝑎 (𝑡)?
𝑑
(a) 𝑥 (𝑡) (b) 𝑥𝑎2 (𝑡)
𝑑𝑡 𝑎

3. A continuous-time signal, given as


𝑥 (𝑡) = 10 cos(400 𝜋𝑡) + 5 cos(800 𝜋𝑡)
is sampled at a rate 600 Hz and then passed through an ideal low-pass filter with unity gain and cut-
off frequency 300 Hz. Determine the output of the filter.

4. A continuous-time signal 𝑥𝑎 (𝑡) is band-limited with 𝑋(𝜔) = 0 for |𝜔| ≥ 𝜔0. If 𝑥𝑎 (𝑡) is sampled
at or above the Nyquist rate, how is the energy in 𝑥[𝑛],
+∞

𝐸𝑑 = ∑ |𝑥[𝑛]|2
𝑛=−∞

related to the energy in 𝑥𝑎 (𝑡),


+∞
𝐸𝑎 = ∫ |𝑥𝑎 (𝑡)|2 𝑑𝑡
−∞

and the sampling period 𝑇𝑠 ?

5. Consider a bandpass signal with center frequency 4𝑊 and bandwidth 𝑊.


(a) Determine the Nyquist rate (minimum sampling rate) for this signal.
(b) Determine the frequencies that are good for sampling the signal so as to achieve alias-free
reconstruction
6. In a sub-band coding scheme for speech, the 0 to 3200 Hz band is split into four bands with
frequency ranges following one of the two designs below.
(a) [0 – 400] [400 – 800] [800 – 2000] [2000 – 3200]
(b) [0 – 800] [800 – 1600] [1600 – 2400] [2400 – 3200]
Determine the minimum overall sampling rates in the above two cases.
Which of these two is a better design? Why?

7. Find the DTFT of each of the following sequences:


1 𝑛 1 |𝑛|
(a) 𝑥 [𝑛] = (2) 𝑢[𝑛 + 3] (b) 𝑥 [𝑛] = (4) (c) 𝑥 [𝑛] = (𝛼 )𝑛 sin(𝑛Ω0 ) 𝑢[𝑛]

8. If 𝑋(Ω) is the DTFT of a sequence 𝑥[𝑛], find the DTFT of


(a) 𝑥 ∗ [−𝑛] (b) 𝑥[2𝑛 + 1]

9. Find the inverse DTFT of


1, 0.25𝜋 ≤ |Ω| ≤ 0.75𝜋
(a) 𝑋(Ω) = { (b) 𝑋(Ω) = cos 2 (Ω)
0, otherwise

10. Let 𝑥[𝑛] be the sequence


𝑥 [𝑛] = 2𝛿 [𝑛 + 2] − 𝛿[𝑛 + 1] + 3𝛿 [𝑛] − 𝛿[𝑛 − 1] + 2𝛿[𝑛 − 2]
Evaluate the following without explicitly finding 𝑋(Ω).
+𝜋
(a) 𝑋(Ω) |Ω=𝜋 (b) ∫−𝜋 |𝑋(Ω)|2 𝑑Ω

11. Find the DTFS representation for the sequence


+∞

𝑥𝑝 [𝑛] = ∑ 𝑥 [𝑛 − 10𝑘 ]
𝑘=−∞

where
1, for 0 ≤ 𝑛 ≤ 5
[ ]
𝑥𝑛 = {
0, otherwise

12. Let 𝑥(𝑡) be a periodic continuous-time signal


𝑥 (𝑡) = 𝐴 cos(200𝜋𝑡) + 𝐵 cos(500𝜋𝑡)
that is sampled at a rate 1 kHz. Find the DTFS coefficients of the sampled signal.
13. Compute the 𝑁-point DFT of the following sequences:
(a) 𝑥 [𝑛] = 𝛼 𝑛 , 0 ≤ 𝑛 ≤ 𝑁
(b) 𝑥 [𝑛] = 𝑢[𝑛] − 𝑢[𝑛 − 𝑛0 ], where 0 < 𝑛0 < 𝑁

14. Find the 10-point inverse DFT of


3, 𝑘=0
𝑋 [𝑘 ] = {
1, 1 ≤ 𝑘 ≤ 9

15. Consider the sequence 𝑥[𝑛] = 4𝛿 [𝑛] + 3𝛿 [𝑛 − 1] + 2𝛿 [𝑛 − 2] + 𝛿[𝑛 − 3]. Let 𝑋[𝑘] be the
6-point DFT of 𝑥[𝑛].
(a) Find the finite length sequence 𝑧[𝑛] that has a 6-point DFT
(b) Find the finite length sequence 𝑞[𝑛] that has a 3-point DFT

16. Let 𝑥[𝑛] and ℎ[𝑛] be two finite-length sequences that are six points long, and let 𝑋[𝑘] and 𝐻[𝑘]
be their respective 8-point DFTs. If 𝑦[𝑛] is the inverse DFT of 𝑌[𝑘] = 𝑋[𝑘]𝐻[𝑘], find the values of
𝑛 for which 𝑦[𝑛] is equal to the linear convolution of 𝑥[𝑛] and ℎ[𝑛].

17. A major problem in recording ECG signal is the appearance of 60 Hz power line interference in the
output. In order to remove this interference, the recorded output analog signal is sampled at Nyquist
sampling rate and then processed in a digital filter described by the difference equation
𝑦[𝑛] = 𝑥 [𝑛] + 𝑎 𝑥[𝑛 − 1] + 𝑏 𝑥 [𝑛 − 2]
Assuming that the analog ECG signal is bandlimited to 1 kHz, determine suitable values for 𝑎 and 𝑏
so that the interference is removed from the recorded signal.

18. Let 𝑋[𝑘] be the 𝑁-point DFT of 𝑥[𝑛]


(a) Find the finite-length sequence that has an 𝑁-point DFT 𝑊𝑁4𝑘 𝑋[𝑘].
(b) Find the finite-length sequence that has an 𝑁-point DFT equal to the real part of 𝑋[𝑘].
(c) Find the 𝑁-point DFT of the real part of 𝑥[𝑛].
(d) Find the 𝑁-point DFT of the imaginary part of 𝑥[𝑛]

19. Consider the sequence 𝑥 [𝑛] = 𝛿[𝑛] + 2𝛿 [𝑛 − 2] + 𝛿[𝑛 − 3].


(a) Find the 4-point DFT of 𝑥 [𝑛].
(b) If 𝑦[𝑛] is the 4-point circular convolution of 𝑥[𝑛] with itself, find 𝑦[𝑛] and its 4-point DFT.
(c) With ℎ[𝑛] = 𝛿 [𝑛] + 𝛿[𝑛 − 1] + 2𝛿 [𝑛 − 3], find the 4-point circular convolution of 𝑥 [𝑛]
with ℎ[𝑛].
20. Compute the periodic convolution of the following two signals that are periodic with period 8.

  3n 
sin  , 0n3 1
n
x[n]    4  ; w[n]    , 0n7
0, 2
 4n7

21. Show how an 𝑁-point FFT program may be used to evaluate


(a) the 𝑁-point DFT of two real-valued sequences.
(b) the 2𝑁-point DFT of a real-valued sequence.
(c) the 𝑁-point inverse DFT of a complex-valued sequence.

22. A complex-valued sequence 𝑥[𝑛] of length 𝑁 = 8192 is to be convolved with a complex-valued


sequence ℎ[𝑛] of length 𝐿 = 512.
(a) Find the number of complex multiplications required to perform this convolution directly.
(b) Find the number of complex multiplications required to perform the convolution using the
overlap-add method with 1024-point radix-2 decimation in time FFTs to evaluate the
convolutions.

23. Speech that is sampled at a rate of 10 kHz is to be processed in real time. Part of the computations
required involve collecting blocks of 1024 speech samples and computing a 1024-point DFT and
a 1024-point inverse DFT. If it takes 1 μs for each real multiply, how much time remains for
processing the data after the DFT and the inverse DFT are computed?

You might also like