Scribd
Upload
19 views5 pages

Chapter2 Solutions

Uploaded by

a1185569145
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
19 views5 pages

Chapter2 Solutions

Uploaded by

a1185569145
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/ 5

1

Chapter 2: Signal Sampling and Reconstruction


Answers

1) According to Nyquist sampling theorem, to avoid aliasing, the minimum sampling rate is
twice of the signal bandwidth:
a) The frequency domain spectrum of x2a (t) is Xa (f ) ∗ Xa (f ), where ∗ sign denotes
convolution. The upper and lower limits of the convolved signal equals the sum of the
upper and lower limits of two participated signals. Therefore, the bandwidth of x2a (t)
is 2B. Based on the Nyquist sampling theorem, the minimum sampling frequency is
4B.
b) According to the scaling property of Fourier transform, F{xa (ct)} = Xa (f /c). As
compared with original signal xa (t), the signal xa (2t) is squeezed in time-domain
and stretched in frequency domain. The bandwidth of xa (2t) is hence 2B and the
minimum sampling frequency is 4B.
c) The Fourier transform of cos(7πBt) is 12 [δ(f − 3.5B) + δ(f + 3.5B)]. Therefore, the
signal xa (t) cos(7πBt) is a bandpass signal, whose frequency response has non-zero
values within the range [2.5B, 4.5B]. According to Nyquist sampling theorem, the
minimum sampling frequency is 2 ∗ fH = 9B. However, for bandpass signal, we
know that the signal can be recovered with under-sampling by a bandpass filter as
long as there is no aliasing after sampling. Here, we have ∆f = 2B and fH = 4.5B.
Since fH is not an integer multiplication of ∆f , we can enlarge the passband to
[2.25B, 4.5B]. Then, the under-sampling rate is 2fH /(fH /∆f ) = 4.5B.

2) No aliasing sampling is achieved with fs ≥ 2fh , where fh is the maximum frequency. For
sinusoid signal, the sampling frequency should satisfy fs > 2fh .
a) fh = 125 Hz, fs > 2fh = 250Hz. To ensure that the resultant discrete-time signal is
periodic, we choose fs = 500Hz and T = 1/fs = 2 × 10−3 , x[n] = x(t|t = nT ) =
A cos(πn/2), which is periodic with period N = 4.
b) fh = 50/π, fs > 2fh = 100/π. To maintain the periodic property, we choose fs =
2

4fh = 200/π, T = 1/fs = π/200, x[n] = A cos(πn/2).


c) f1 = 50Hz, f2 = 80Hz, f3 = 180Hz, and x(t) is periodic. To maintain the periodic
property, we let fs = 4fh = 720Hz. x[n] = cos(5πn/36) + cos(2πn/9) + cos(πn/2).
x[n] is periodic with period N = 72.
d) It is a sinc function with fh = 200Hz. If fs = 2fh = 400Hz, we have x[n] = x(t|t =
10 sin(2πn/3)
n/400) = 0. Hence, we may choose fs = 3fh = 600Hz, x[n] = 2πn/3
.

3) It is a bandpass signal with ∆f = 25 Hz and fH = 50 Hz. When sampling frequency


fs = 2fH = 100 Hz, there is no aliasing according to Nyquist sampling theorem. When
fs = 2fH /(fH /∆f ) = 50 Hz, there is no aliasing since it satisfies the under-sampling
conditions. There will be aliasing if the sampling rate is 25 Hz. The frequency spectrum
of sampled signal is shown in Fig. 1.

Fig. 1

4) a) This signal is passed by a A/D converter with sampling rate Ωs = 2π/T1 and then
3

down-sampled by an integer factor M . The effective sampling rate is Ω̄s = 2π/M T1 .


According the Nyquist sampling theorem, for signal recovering, we must have Ω̄s ≥
2ΩH , i.e., M T1 ≤ 0.25 ms.
b) Since M T1 ≤ 0.25 ms, M ≤ 0.25 × 30. The maximum value of integer M is 7.

5) Note that ω = ΩT . The complete frequency response will repeat every 2π in the plot.
After decimation (reducing the sampling rate by an integer factor), the effective sampling
rate is reduced. The amplitude of the frequency reduces and the frequency response is
stretched. The frequency response is shown in Fig 2.

Fig. 2

6) After interpolation, the frequency response will be squeezed and the amplitude will in-
creases by the factor L. The frequency response for various L is shown in Fig. 4;

1, |ω| ≤ ω1

jω 3
7) X(e ) =
0, ω1 < |ω| ≤ π

3 3
4

Fig. 3

3 sin( 53 ω1 n)
a) y[n] = 5
πn
3
1
b) y[n] = 5
δ[n].
5

Fig. 4

You might also like