SIGNALS AND SYSTEMS
Academic year 2024-2025. Instructor: Dr. Marco Fabris.
Solved exercises - Part 2
1. Compute the generalized derivative of the continuous-time signal
x(t) = sign(t)e(2+j)t , t∈R
where the “sign” signal is defined as:
1,
if t > 0
sign(t) = 0, if t = 0
−1, if t < 0
Solution. Formally, recalling the product rule for differentiation and the property of the
Dirac unit impulse δ according to which f (t)δ(t) = f (0)δ(t) for any function f continuous
at t = 0, it holds that
" #
d d d
x(t) = sign(t) e(2+j)t + sign(t) e(2+j)t
dt dt dt
= 2δ(t)e(2+j)t + sign(t)(2 + j)e(2+j)t
= 2δ(t) + sign(t)(2 + j)e(2+j)t .
The above result is a consequence of the fact that the signal sign(t) is constant, except
for a discontinuity of amplitude 2 at t = 0.
In particular, we obtain
d d
Re x(t) = Re x(t) = 2δ(t) + sign(t)e2t (2 cos t − sin t)
dt dt
= 2δ(t) + e2t 2 sign(t) cos t − sin |t|
d d
Im x(t) = Im x(t) = sign(t)e2t (cos t + 2 sin t)
dt dt
= e2t sign(t) cos t + 2 sin |t|
2. Compute the generalized derivative of the continuous-time signal
1
x(t) = sign(t) − 1l(t) + t2 1l(t), t ∈ R.
2
Solution. Note that sign(t) = 1l(t) − 1l(−t). The signal x(t) can therefore be written as
x(t) = 12 (1l(t) − 1l(−t)) − 1l(t) + t2 1l(t) = − 21 (1l(t) + 1l(−t)) + t2 1l(t) = − 12 + t2 1l(t).
The generalized derivative is dtd x(t) = 2t1l(t) + t2 δ(t) = 2t1l(t).
3. Consider the signal x(t), t ∈ R, periodic with period T = 2, defined for t ∈ [0, 2) as:
(
3t, 0 ≤ t < 1,
x(t) =
0, 1 ≤ t < 2.
d
a) Sketch the plot of x(t). b) Compute the generalized derivative y(t) = dt
x(t), t ∈ R.
Solution
a.)
x(t) 6
3
2
1
-
−4 −3 −2 −1 1 2 3 4 t
b.) The plot of the generalized derivative of x can be sketched by inspection.
6
d
dt
x(t)
-
−4 −3 −2 −1 1 2 3 4 t
−1
−2
? ? −3 ? ?
To formally compute the generalized derivative, we write
x(t) = ∞ k=−∞ 3(t − 2k)(1l(t − 2k) − 1l(t − 1 − 2k))
P
(i.e., x(t) is the periodic repetition with period T = 2 of the given signal in the interval
[0, 2)). The generalized derivative is then yielded by
dx
= ∞ k=−∞ 3(1l(t − 2k) − 1l(t − 1 − 2k)) + 3(t − 2k)(δ(t − 2k) − δ(t − 1 − 2k))
P
dt
applying the rules of the Dirac δ function, we find
dx
= ∞ k=−∞ 3(1l(t − 2k) − 1l(t − 1 − 2k)) − 3δ(t − 1 − 2k).
P
dt
4. For the discrete-time system described by the relation
n+10
X
y(n) = x(k),
k=n−10
verify whether the following properties are satisfied: a. causality, b. BIBO-stability.
Solution. a. y(n) depends on both past and future values of x(·) relative to n. The
system is not causal.
b. If x(·) is a bounded signal, say |x(n)| ≤ B < ∞, ∀ n, then |y(n)| = | n+10
k=n−10 x(k)| ≤
P
Pn+10
k=n−10 |x(k)| ≤ (n + 10 − (n − 10) + 1)B = 21B < ∞. The system is BIBO stable.
5. For each of the following systems:
(a) y(t) = x(t − 2), t ∈ R,
n−1
3k x(k),
X
(b) y(n) = n ∈ Z,
k=−∞
(c) y(t) = x(0)2 − t, t ∈ R,
verify whether the following properties hold: a. causality, b. linearity, c. time invariance,
d. BIBO stability.
Solution.
(a) The relation between input and output is a simple delay translation by T = 2.
Hence, it is a a. causal, b. linear, c. time-invariant, d. BIBO stable system.
(b) The output y(n) is obtained as a linear combination of past samples x(k), k < n,
of the input, but the coefficients are not a function of the difference n − k alone.
Moreover, the response to the bounded input x(n) = u(n) is the unbounded signal
y(n) = 12 (3n −1) 1l(n). Thus, this system is a. causal, b. linear, c. not time-invariant,
d. not BIBO stable.
(c) The output y(t) solely depends on (the time t and) the input x(0), which can be
from the past, present, or future, depending on the value of t ∈ R. Moreover, the
response to the zero input x(t) ≡ 0 is the non-zero and unbounded signal y(t) = −t.
Finally, while the zero input satisfies the property of invariance x(t) = x(t − T ) for
every T ∈ R, the corresponding output can be written as y(t) = y(t − T ) only for
T = 0. Consequently, this is a a. non-causal, b. non-linear, c. not time-invariant,
d. not BIBO stable system.
6. For each of the following systems:
(a) y(t) = x(t + 5) x(t − 1), t ∈ R,
Z t
(b) y(t) = x(s) ds, t ∈ R,
−∞
(c) y(t) = cos(t − 2) x(t), t ∈ R,
verify whether the following properties hold: a. causality, b. linearity, c. time invariance,
d. BIBO-stability.
Solution.
(a) It is not causal because y(t) depends on x(t + 5). It is not linear. It is time-invariant
since the input x(t) = x(t − T ) produces the output y(t) = x(t + 5) x(t − 1) =
x(t − T + 5) x(t − T − 1), which equals y(t − T ). It is BIBO stable because if
|x(t)| ≤ B then |y(t)| ≤ B 2 .
(b) It is causal because y(t) solely depends on {x(τ ), −∞ < τ ≤ t}. It is linear. It
is time-invariant
Rt Rt
because applyingR input x(t) = x(t − T ) results in output y(t) =
t−T ′ ′
−∞ x(τ )dτ = −∞ x(τ − T )dτ = −∞ x(τ )dτ = y(t − T ). It is not BIBO stable;
one can verify this by considering the response to a constant input.
(c) It is causal. It is linear. It is not time-invariant. It is BIBO stable since |y(t)| ≤
|x(t)|.
7. For the continuous-time system
y(t) = t x(t − 2) − cos(t + 2) x(t)
verify whether the following properties hold: a. causality, b. linearity, c. time invariance,
d. BIBO stability.
Solution. a. Since the output y at time t depends only on the input x(s) at times s = t
and s = t − 2, which are either previous or equal to t, the system is causal.
b. The presence of the quadratic term x(t−2)x(t) in the input-output relationship makes
the system nonlinear. Although y(t) ≡ 0 if x(t) ≡ 0, neither homogeneity nor additivity
is satisfied. For example, for the input x1 (t) = αx(t), the output is y1 (t) = α αt x(t −
2) − cos(t + 2) x(t), which is generally different from αy(t) = α t x(t − 2) − cos(t + 2) x(t).
c. The time variability of the coefficients t and − cos(t + 2) is the reason why the system
is not time-invariant.
the input: x1 (t) = x(t − t0 ), results in the
Specifically, translating
output y1 (t) = t x(t − t0 − 2) − cos(t + 2) x(t − t0 ), which is generally different from
y(t − t0 ) = (t − t0 ) x(t − t0 − 2) − cos(t − t0 + 2) x(t − t0 ).
d. The system is not even BIBO stable due to the unbounded nature of the coefficient t.
For a bounded input x(t) ≡ 1, the output is unbounded: y(t) = t − cos(t + 2).
8. For each of the following systems:
(a) y(n) = max{|x(n)|, |x(n − 1)|}, n ∈ Z,
Rt
(b) y(t) = −∞ e−3(t−τ ) x(τ ) dτ , t ∈ R,
R 2t 2
(c) y(t) = −∞ |t − τ | x(τ ) dτ , t∈R
verify whether the following properties hold:
a. causality, b. linearity, c. time invariance, d. BIBO stability.
Solution. System (a) is causal, nonlinear, time-invariant, and BIBO stable. System (b)
is causal, linear, time-invariant, and BIBO stable, with impulse response h(t) = e−3t u(t).
Finally, system (c) is not causal, is linear, not time-invariant, nor BIBO stable.