Professional Documents
Culture Documents
Ecen 644 - Advanced Digital Signal Processing
Ecen 644 - Advanced Digital Signal Processing
http://www.comm.utoronto.ca/~dkundur/course/discrete-time-systems/
PROCESSING
HOMEWORK #1 - SOLUTIONS
1.2
For a sinusoid of the form x(n ) = A cos(ω ⋅ n + θ ) , the frequency can be expressed as
ω
f = .
2π
(a) cos(0.01π ⋅ n )
ω 0.01π 1
f = = = is rational and (1,200) are relatively prime.
2π 2π 200
30
(b) cos π ⋅ n
105
ω 30π 1 1
f = = ⋅ = is rational and (1,7) are relatively prime.
2π 105 2π 7
30
cos π ⋅ n is periodic with fundamental period N = 7.
105
(c) cos(3π ⋅ n )
ω 3π 3
f = = = is rational and (3,2) are relatively prime.
2π 2π 2
(d) sin (3 ⋅ n )
ω 3 3
f = = = is not rational.
2π 2π 2π
sin (3 ⋅ n ) is non-periodic.
62
(d) sin π ⋅ n
10
ω 62π 1 31
f = = ⋅ = is rational and (31,10) are relatively prime.
2π 10 2π 10
62
sin π ⋅ n is periodic with fundamental period N = 10.
10
1.4
(a) As mentioned previously the fundamental period of a sinusoid when it exists is equal to
ω k
N where f 0 = = and k, N are relatively prime.
2π N
2π ⋅ kn ω k
For s k (n ) = exp j , k = 0,1,2,... , we have f = = .
N 2π N
Suppose now that k and N are not relatively prime, i.e. ∃α such that α = GCD (k , N ) and
k = k '⋅α and N = N '⋅α where k’ and N’ are relatively prime. We can now rewrite:
ω k k '⋅α k'
f0 = = = = .
2π N N '⋅α N '
N 7 7 7 7 7 7 7 7
k 0 1 2 3 4 5 6 7
GCD(k,N) 7 1 1 1 1 1 1 7
Np 1 7 7 7 7 7 7 1
N 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16
k 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
GCD(k,N) 16 1 2 1 4 1 2 1 8 1 2 1 4 1 2 1 16
Np 1 16 8 16 4 16 8 16 2 16 8 16 4 16 8 16 1
1.6
1
Because x(n ) is a discrete-time sinusoid obtained by sampling at a rate Fs = a
T
1
continuous-time sinusoid x a (t ) with fundamental period Tp = , we find x(n ) as follows:
F0
x a (t ) = A cos(2πF0 t + θ ) x(n ) = x a (nT ) = A cos(2π ⋅ F0 nT + θ )
F0
= A cos 2π ⋅ n +θ
Fs
T
= A cos 2π ⋅ n +θ
Tp
T T k
(a) If = f is rational i.e., = , we have straight from definition B1 from the
Tp Tp N
textbook that x(n ) is periodic.
k
(b) x(n ) is periodic implies f rational i.e., f = . N here represents the period in
N
samples which means that to get the period in seconds N must be multiplied by T.
k T
Then we get Td = N ⋅ T = T but f = and therefore the final result:
f Tp
Tp
Td = k T = kT p
T
1.7
(a) To avoid any aliasing problem, it is known that a signal must be sampled as at least
twice its maximum frequency : Fs ≥ 2 ⋅ Fmax . In this case Fmax = 10kHz and
therefore the sample frequency should be at least 20kHz.
Fs ≥ 20kHz
F1 5 3
x(n ) = cos 2π ⋅ n = cos(2π ⋅ f ⋅ n ) = cos π n or cos π n
Fs 4 4
From the equality F0 = f ⋅ Fs , we obtain F0 = 3 kHz and we can therefore conclude that
for Fs = 8 kHz, the frequency F1 = 5 kHz is aliased to 3 kHz.
2.2
Original Signal:
x(n)-- 1/2
0
-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(a)
1
x(n-2)
-- 1/2
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(b)
1
x(4-n)
-- 1/2
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(c)
x(n+2) 1
-- 1/2
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(d)
1
x(n)u(2-n)
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(e)
x(n-1)d(n-3) 1
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(f)
1
x(n2)
-- 1/2
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(g)
x e(n)=[x(n)+x(-n)]/2 1
-- 1/2
-- 1/4
0
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
(h)
-- 1/2
x o(n)=[x(n)-x(-n)]/2
-- 1/4
-- -1/4
-- -1/2
-1
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
n
2.4
The even and odd parts of a signal x(n ) can be found using the following formulas:
xeven (n ) =
[x(n ) + x(− n )] xodd (n ) =
[x(n ) − x(− n )]
2 2
Where xeven (n ) being the even part verifies xeven (− n ) = xeven (n ) and xodd (n ) being the odd
part verifies xodd (− n ) = − xodd (n ) .
Therefore any x(n ) can be written as x(n ) = xeven (n ) + xodd (n ) i.e., any signal can be
decomposed into an even and an odd component and this decomposition is unique.
Example:
{ }
x(n ) = 2 3 4 5 6
↑
x(− n ) = {6 5 4 3 2}
↑
x (n ) = {4 4 4 4 4}
even
↑
x (n ) = {− 2 − 1 0 1 2}
odd
↑
2.5
∞ ∞
2⋅ xeven (n ) ⋅ x odd (n ) =2 ⋅ xeven (− m ) ⋅ xodd (− m )
n = −∞ m = −∞
We know that xeven (n ) being the even part verifies xeven (− n ) = x even (n ) and xodd (n ) being
the odd part verifies xodd (− n ) = − x odd (n ) , therefore we can write:
∞ ∞
2⋅ xeven (− m ) ⋅ x odd (− m ) = 2 ⋅ xeven (m ) ⋅ (− x odd (m ))
m = −∞ m = −∞
∞
= − 2⋅ xeven (n ) ⋅ xodd (n )
n = −∞
∞ ∞
We then obtain the equality 2 ⋅ xeven (n ) ⋅ xodd (n ) = − 2 ⋅ xeven (n ) ⋅ x odd (n ) which proves
n = −∞ n = −∞
∞
that 2 ⋅ xeven (n ) ⋅ xodd (n ) = 0 and finally:
n = −∞
∞ ∞ ∞
E= x 2 (n ) = 2
x even (n ) + 2
xodd (n )
n = −∞ n = −∞ n = −∞
2.7
Recall that:
- A system is static if the output at time n depends only on inputs at the same time n.
F [a1 x1 (n ) + a 2 x 2 (n )] = a1 F [x1 (n )] + a 2 F [x 2 (n )]
- A system is (BIBO) stable if when the input x(n ) is bounded, the output y (n ) is also
bounded.
(a) y (n ) = cos[x(n )]
- the output at time n depends only on the input at time n, the system is static.
The system y (n ) = cos[x(n )] is static, nonlinear, time invariant, causal and stable.
(g) y (n ) = x(n )
- the output at time n depends only on the input at time n, the system is static.
The system y (n ) = x(n ) is static, nonlinear, time invariant, causal and stable.
(j) y (n ) = x(2n )
- the output at time n depends on inputs at time future time 2n, the system is dynamic.
The system y (n ) = x(2n ) is dynamic, linear, time variant, noncausal and stable.
x(n ) if x(n ) ≥ 0
(k) y (n ) =
0 if x(n ) < 0
- the output at time n depends only on the input at time n, the system is static.
x(n ) if x(n ) ≥ 0
The system y (n ) = is static, nonlinear, time invariant, causal and
0 if x(n ) < 0
stable.