x[n] in terms of Sc ( j).
EEE4001F: Problem Set 3
(b) We want to simulate the multipath system with a discrete-time system
by choosing H (e j ω ) in the figure so that the output r [n] = x c (nT )
1. The sequence
π when the input is s[n] = sc (nT ). Determine H (e j ω ) in terms of T and
x[n] = cos
4
n , −∞ < n < ∞ PSfragτdreplacements
.
(c) Determine the impulse response h[n] in
was obtained by sampling a continuous-time signal
x c (t) = cos(0 t), −∞ < t < ∞, H (e j ω )
s[n] = sc (nT ) r [n] = x c (nT )
at a sampling rate of 1000 samples/s. What are two possible values of 0 when (i) τd = T and (ii) τd = T /2.
that could have resulted in the sequence x[n]?
4. In the figure below,
2. The continuous-time signal PSfrag replacements
Discrete−time
x c (t) = sin(20π t) + cos(40π t) C/D D/C
x c (t) system yr (t)
x[n] y[n]
is sampled with a sampling period T to obtain the discrete-time signal
T T
πn
2π n
x[n] = sin + cos .
5 5 assume that
(a) Determine a choice of T consistent with this information. H (e j ω ) = jω/T , −π ≤ ω < π,
(b) Is your choice for T in part (a) unique? If so, explain why. If not, and T = 1/10 seconds.
specify another choice of T consistent with the information given. (a) For each of the following inputs x c (t), find the corresponding output
3. A simple model of a multipath communication channel is indicated below:
PSfrag replacements yc (t).
i. x c (t) = cos(6π t)
sc (t)
ii. x c (t) = cos(14π t)
α Delay x c (t) = sc (t) + αsc (t − τd ) (b) Are the outputs yc (t) those you would expect from a differentiator?
τd
5. Which of the following signals can be downsampled by a factor of 2 using
Assume that sc (t) is bandlimited such that Sc ( j) = 0 for || ≥ π/T the system
PSfrag replacements
and that x c (t) is sampled with a sampling period T to obtain the sequence
Compressor
x[n] = x c (nT ).
↓M
(a) Determine the Fourier transform of x c (t) and the Fourier transform of x[n] x d [n] = x[n M]
1 2
without any loss of information? PSfrag replacements
C/D D/C
(a) x[n] = δ[n − n 0 ], for n 0 some unknown integer
PSfrag replacements x c (t) x[n] xr (t)
(b) x[n] = cos(π n/4)
T T
(c) x[n] = cos(π n/4) + cos(3π n/4) PSfrag replacements
Determine the range of values for T for which x r (t) = x c (t).
(d) x[n] = sin(π n/3)/(π n/3).
(e) x[n] = (−1)n sin(π n/3)/(π n/3). 8. A continuous-time signal x c (t), with Fourier transform
X c ( j)
6. Consider the system 1
H (e j ω ) 1
↓3 ↑3 −0 − 20 0 0
2
0
x[n] x d [n] x e [n] xr [n]
− π3 π
3 is sampled with sampling period T = 2π/ 0 to form the sequence
x[n] = x c (nT ).
For each of the following input signals x[n], indicate whether the output
(a) Sketch the Fourier transform X (e j ω ) for |ω| < π .
xr [n] = x[n].
(b) The signal x[n] is to be transmitted across a digital channel. At the
(a) x[n] = cos(π n/4)
receiver, the original signal x c (t) must be recovered. Draw a block
(b) x[n] = cos(π n/2)
diagram of the recovery system and specify its characteristics. Assume
(c) that ideal filters are available.
2
sin(π n/8)
x[n] = (c) In terms of 0 , for what range of values of T can x c (t) be recovered
πn
from x[n]?
PSfrag replacements
Hint: Use the modulation property of the Fourier transform to find
PSfrag replacements
9. In the system
X (e j ω ).
7. The continuous-time signal x c (t) with the Fourier transform C/D H (e j ω ) D/C
x c (t) yc (t)
X c ( j)
1
T1 T2
X c ( j) and H (e j ω ) are as shown below:
−0 − 23 0 0 2
3 0
0
is passed through the system
3 4
PSfrag replacements
X c ( j) 11. Consider the complex sequence
1
e j ω0 n , 0≤ n ≤ N −1
x[n] =
0, otherwise.
−2π × 5 × 103 2π × 5 × 103
1 H (e j ω ) (a) Find the Fourier transform X (e j ω ) of x[n].
(b) Find the N -point DFT X [k] of the finite-length sequence x[n].
ω (c) Find the DFT of x[n] for the case ω0 = 2π k0 /N , where k0 is an
−π − π2 π
2 π
integer.
Sketch and label the Fourier transform of yc (t) for each of the following 12. The two 8-point sequences x 1 [n] and x 2 [n] shown below have DFTs
cases: X 1 [k] and X 2 [k] respectively:
(a) 1/T1 = 1/T2 = 104 . c
b d x 1 [n]
(b) 1/T1 = 1/T2 = 2 × 104 .
a e
(c) 1/T1 = 2 × 104 , 1/T2 = 104 . n
0 1 2 3 4 5 6 7 8
(d) 1/T1 = 104 , 1/T2 = 2 × 104 .
c
PSfrag replacements d b x 2 [n]
10. Compute the DFT of each of the following finite-length sequences
e a
considered to be of length N (where N is even):
n
(a) x[n] = δ[n] 0 1 2 3 4 5 6 7 8
(b) x[n] = δ[n − n 0 ], 0 ≤ n 0 ≤ N − 1 Determine the relationship between X 1 [k] and X 2 [k].
1, n even, 0≤n ≤ N −1 13. Consider the following finite-length sequence x[n]:
(c) x[n] =
0, n odd, 0≤n ≤ N −1 2 2
1 1
1, 0 ≤ n ≤ N /2 − 1 n
(d) x[n] = −2 −1 0 1 2 3 4 5 6 7
0, N /2 ≤ n ≤ N − 1
The five-point DFT of x[n] is denoted by X [k]. Plot the sequence y[n]
a n , 0≤n ≤ N −1 whose DFT is
(e) x[n] =
0, otherwise. Y [k] = W5−2k X [k].
14. Two finite-length sequences are shown below:
5 6
2 (b) Sketch the finite-length sequence w[n] whose six-point DFT is
1 x 1 [n]
W [k] = Re{X [k]}.
−1 0 1 2 3 4 5 6 7 8 9 10 (c) Sketch the finite-length sequence q[n] whose three-point DFT is
3
2 Q[k] = X [2k], k = 0, 1, 2.
PSfrag replacements 1 x 2 [n]
17. Consider the following finite-length sequence:
6
−1 0 1 2 3 4 5 6 7 8 9 10 5 x[n]
PSfrag replacements 4
3
Assume that x 1 [n] and x 2 [n] are zero outside of the region shown in the
figure. Let x 3 [n] be the 8-point circular convolution of x 1 [n] with x 2 [n]. n
Determine x 3 [2]. Sketch the sequences
15. The following figure shows two finite-length sequences x 1 [n] and x 2 [n]. x 1 [n] = x[((n − 2))4 ], 0 ≤ n ≤ 3,
3
x 1 [n] 2 x 2 [n] and
1 1 x 2 [n] = x[((−n))4 ], 0 ≤ n ≤ 3.
PSfrag replacements 1 2 4
n n
−1 0 3 4 5 −1 0 1 2 3 5 6 7 18. Sketch the N -point circular convolution of the following two sequences
−1 −1
6
−2 PSfrag replacements 5 x 1 [n] x 2 [n]
4 1 1
3
What is the smallest N such that the N -point circular convolution of x 1 [n] 2
1
and x 2 [n] is equal to the linear convolution of these sequences? n n
0 1 2 3 4 5 0 1 2 3 4
16. Consider the real finite-length sequence
for N = 6 and N = 10.
4
3 x[n]
2
PSfrag replacements 1
n
−2 −1 0 1 2 3 4 5 6
(a) Sketch the finite-length sequence y[n] whose six-point DFT is
Y [k] = W64k X [k],
where X [k] is the six-point DFT of x[n].
7 8