Signals and Systems 3102:
Tutorial: Fourier Transform of DT signals
Feb 2025
Q1) For Q1a,b,c) determine the period (if any) of the following discrete sinusoid signals:
a) 𝑥𝑥1 [𝑛𝑛] = 𝐴𝐴 𝑐𝑐𝑐𝑐𝑐𝑐( 𝜋𝜋𝜋𝜋/6 + 𝜋𝜋/3)
b) 𝑥𝑥2 [𝑛𝑛] = 𝑠𝑠𝑠𝑠𝑠𝑠( 1.8𝜋𝜋𝜋𝜋)
c) 𝑥𝑥3 [𝑛𝑛] = 𝑐𝑐𝑐𝑐𝑐𝑐( 0.5𝑛𝑛)
d) Write a python/Matlab script to plot the above 3 signals for n=-20 to 20.
e) Plot 𝑥𝑥𝐶𝐶𝐶𝐶 (𝑡𝑡) = sin (ω𝑡𝑡) with t from -10 to 10 in step of 1/1000 seconds and
𝑥𝑥𝐷𝐷𝐷𝐷 [𝑛𝑛] = sin (Ω𝑛𝑛), given Ω = 𝜔𝜔 = 1.8𝜋𝜋. Write a python/Matlab script to plot the
figures, comment on the difference between 𝑥𝑥𝐷𝐷𝐷𝐷 [𝑛𝑛] 𝑣𝑣𝑣𝑣 𝑥𝑥𝐶𝐶𝐶𝐶 (𝑡𝑡) .
f) With respect to Q1e, what are the units of Ω and 𝜔𝜔 respectively, and comment on their
difference.
g) Express 𝑥𝑥[𝑛𝑛] = 3 cos(Ω𝑛𝑛 + 𝜙𝜙) as 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴(Ω𝑛𝑛) + 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵(Ω𝑛𝑛), i.e, a sum of cosine and
𝜋𝜋 𝑟𝑟𝑟𝑟𝑟𝑟 𝜋𝜋
sine wave of the same frequency Ω = 6
(𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠) and 𝜙𝜙 = 3 . In other words, find A
and B. Hint: use the trigo identity cos(X+Y).
Q2) Fourier Transform representations of pure sinusoids:
Determine the Fourier transforms of the following discrete time signals
𝜋𝜋
i) 𝑥𝑥[𝑛𝑛] = cos �4 𝑛𝑛�
𝜋𝜋 𝜋𝜋
ii) 𝑥𝑥[𝑛𝑛] = cos �3 𝑛𝑛� + sin ( 4 𝑛𝑛)
a) Find the DTFS of above by using the Euler representation of sinusoids. Establish the
appropriate N, and interpret value of k for the spectra.
𝜋𝜋 𝜋𝜋
b) For 𝑥𝑥[𝑛𝑛] = cos �3 𝑛𝑛� + sin ( 4 𝑛𝑛)
(i) write a matlab/or Python code to generate x[n] for 1 period, i.e Find N. Then use
DFT function to calculate X[k].
(ii) From the DFT values, sketch the magnitude and phase spectra of the corresponding
DFT, DTFS and DTFT representations. Use the appropriate horizontal axis labels,
and also the type of ‘stem’ arrowhead for the values
Q2c(i) Find x[n] given the following DFT representation of the signal, choose the appropriate N,
a) 𝑋𝑋[𝑘𝑘] = [4, 3 + 𝑗𝑗3, 2, 3 – 𝑗𝑗3]𝑇𝑇
−𝑗𝑗𝑗𝑗 𝑗𝑗𝑗𝑗
b) 𝑋𝑋[𝑘𝑘] = [ 8,0,16𝑒𝑒 2 , 0,0,0, 16𝑒𝑒 2 , 0]
Q2c(ii) a) how are the DTFS and DFT equations related?
b) Discuss key considerations when applying a DFT on
an extracted N-sampled sequence 𝑥𝑥[𝑛𝑛].
Q3) Find the discrete-time Fourier Transform (DTFT) representation of the following sequence,
and plot the magnitude and phase spectra. Use different lengths of h[n], e.g, 2,5,20 to view the
approximation to 𝑋𝑋(Ω) via the DTFT analysis equation by DFT.
1 𝑛𝑛
ℎ[𝑛𝑛] = �2� 𝑢𝑢[𝑛𝑛]
𝜋𝜋 𝜋𝜋
Q4) Given 𝑥𝑥[𝑛𝑛] = 2 cos � 𝑛𝑛 + � being processed by the following LTI system with impulse
20 3
response
a) ℎ[𝑛𝑛] = {1,2, −3,0,0, … }, where the elements are index at time n=0,1,2,3,4,…
1 0 1 1 1 2 1 3 1 𝑛𝑛
b) ℎ2 [𝑛𝑛] = �� � , � � , � � , � � , … � = � � 𝑢𝑢[𝑛𝑛],
2 2 2 2 2
i) Method 1: Find closed form solution to find 𝐻𝐻(Ω) to scale the eigen function to
find the steady state response of 𝑦𝑦[𝑛𝑛]
ii) Method 2: Use sum of scaled delayed version of 𝑦𝑦[𝑛𝑛] = 𝑥𝑥[𝑛𝑛] ∗ ℎ[𝑛𝑛]
iii) Method 3: Use matlab conv() function with the impulse response ℎ[𝑛𝑛] (use L
samples of ℎ[𝑛𝑛]) to evaluate 𝑦𝑦[𝑛𝑛]
iv) Method 4: Use matlab’s y= filter(B,A,x)
Plot these solutions and compare/comment, especially discuss about the first
beginning N samples (N = 0,1,2…)
Q5) a) What are the purposes of zero padding?
b) Given a LTI system with impulse response h[n] = [1,2,3,4] for index n=0,1,2,3 and
value 0 otherwise, and input x[n] = [1,-2] for index n = 0,1 and value 0 otherwise.
i) Find y[n] = x[n]*h[n] using the convolution equation by hand.
ii) Find y[n] using fft approach to show that your results are the same.
----------------------------- End of Tutorial -------------------------------